|
|
||||||||||
|
J Am Coll Cardiol, 2005; 45:1064-1071, doi:10.1016/j.jacc.2004.12.069 © 2005 by the American College of Cardiology Foundation |




* Echocardiographic Core Laboratory, Baylor University Medical Center, Dallas, Texas
Mayo Clinic Scottsdale, Scottsdale, Arizona
University of California, San Diego, San Diego, California
University of Colorado, Denver, Colorado
|| Mayo Clinic Rochester, Rochester, Minnesota
¶ National Heart, Lung, and Blood Institute, Washington, DC
# University of Wisconsin, Madison, Wisconsin
** University of Pennsylvania, Philadelphia, Pennsylvania
Manuscript received October 11, 2004; revised manuscript received December 3, 2004, accepted December 20, 2004.
* Reprint requests and correspondence: Dr. Paul A. Grayburn, Baylor Heart and Vascular Institute, Baylor University Medical Center, 621 North Hall Street, Dallas, Texas 75226 (Email: paulgr{at}baylorhealth.edu).
| Abstract |
|---|
|
|
|---|
BACKGROUND: Previous studies indicate that echocardiographic measurements of LV size and function, mitral deceleration time, and mitral regurgitation (MR) predict adverse outcomes in HF. However, complete quantitative echocardiograms evaluating all of these parameters have not been reported in a prospective randomized clinical trial in the era of modern HF therapy.
METHODS: Complete echocardiograms were performed in 336 patients at 26 sites and analyzed by a core laboratory. A Cox proportional-hazards regression model was used to determine which echocardiographic variables predicted the primary end point of death or the secondary end point of death, HF hospitalization, or transplant. Significant variables were then entered into a multivariable model adjusted for clinical and demographic covariates.
RESULTS: On multivariable analysis adjusted for clinical covariates, only LV end-diastolic volume index predicted death (events = 75), with a cut point of 120 ml/m2. Three echocardiographic variables predicted the combined end point of death (events = 75), HF hospitalization (events = 97), and transplant (events = 9): LV end-diastolic volume index, mitral deceleration time, and the vena contracta width of MR. Optimal cut points for these variables were 120 ml/m2, 150 ms, and 0.4 cm, respectively.
CONCLUSIONS: Echocardiographic predictors of outcome in advanced HF include LV end-diastolic volume index, mitral deceleration time, and vena contracta width. These variables indicate that LV remodeling, increased LV stiffness, and MR are independent predictors of outcome in patients with advanced HF.
| |||||||||||||
| Methods |
|---|
|
|
|---|
35% were randomly assigned to placebo or bucindolol therapy, with mortality as a primary end point. Patients were excluded if they had HF due to a reversible cause, uncorrected primary valvular disease, hypertrophic cardiomyopathy, untreated thyroid disease, pericardial disease, amyloidosis, active myocarditis, prosthetic valve dysfunction, or recent myocardial infarction (<6 months). All patients were at least 18 years old and provided written, informed consent for the main trial and for the echocardiographic substudy. Complete two-dimensional and Doppler echocardiograms were performed in 355 patients enrolled in the BEST study at 26 clinical sites. Sites were selected on the basis of experience in echocardiographic clinical research, interest in this substudy, and submission of three protocol-specific qualifying echocardiograms. All BEST participants at these sites were asked to participate in the echocardiographic substudy. Echocardiograms of poor technical quality were excluded in 19 patients (5%); the remaining 336 patients comprise this report.
Echocardiography. Blood pressure was taken by cuff at the time of echocardiography. A thorough two-dimensional and Doppler echocardiographic study was then performed according to a specific imaging protocol. Each site received a training manual and videotape detailing the specific views and measurements required to assess LV systolic and diastolic function, LV geometry, LV mass, and MR. All echocardiographic data were recorded on S-VHS videotape labeled with the date, BEST study patient identification number, and patient initials, and sent to the core laboratory for analysis and quality control. Studies were interpreted without knowledge of clinical characteristics or treatment assignment.
Analysis of LV systolic function and geometry. At each participating site, sonographers were instructed to adjust gain settings to optimize visualization of the LV endocardial contours while avoiding excessive gain artifact. Standard parasternal, apical, and subcostal views were obtained. Specific instruction included placing the transducer as far laterally and caudally as possible in the apical windows to maximize LV cavity size and avoid foreshortening. Technically difficult studies were excluded by the core laboratory when inadequate endocardial border visualization in the apical views precluded measurement of LV volumes. Left ventricular end-systolic and end-diastolic dimensions, posterior wall thickness, septal wall thickness, and left atrial dimension were measured according American Society of Echocardiography recommendations (12). The LV volumes and LVEF were quantified at the core laboratory using the biplane Simpson's rule (12). Stroke volume and cardiac output were determined by pulsed Doppler technique as recommended by the American Society of Echocardiography (13). The LV geometry was assessed by a ratio of the major-to-minor axis at end-diastole (14) and as the sphericity index, which is the ratio of the LV end-diastolic volume to a sphere whose radius equals the length of the LV (15). The LV mass was assessed by the 5/6 area-length method as recommended by the American Society of Echocardiography (12). Circumferential wall stress was calculated from cuff systolic blood pressure and end-systolic echocardiographic measurements (16).
Analysis of LV diastolic function. Left ventricular filling patterns were determined by pulsed Doppler technique with a 1- to 2-mm sample volume positioned between the tips of the mitral leaflets parallel to mitral inflow from the four-chamber view (8,9). Mitral inflow patterns were analyzed for maximal E and A velocities, E/A ratio, A-wave duration, and deceleration time. Deceleration time was calculated as the time between the peak E-wave and its deceleration slope extrapolated to the zero baseline. By slightly repositioning the sample volume to overlap mitral inflow and aortic outflow, isovolumic relaxation time was assessed as the time from the end of aortic ejection to the start of mitral inflow. Pulmonary venous flow velocities were obtained with a 2- to 4-mm sample volume positioned 1 to 2 cm into the right and left upper pulmonary veins from a modified four-chamber view (13). Peak systolic and diastolic flow velocities were recorded. In addition, the peak A-wave reversal and A duration were measured.
Assessment of MR. Doppler color flow mapping was used to identify the presence or absence of MR. Gain settings were optimized by reducing the gain to the point where background noise disappeared. Frame rate was maximized by minimizing the sector angle and depth settings to allow visualization of the entire contour of the left atrium. The left atrium was interrogated from multiple acoustic windows using adjustments in transducer angulation to identify the largest MR velocity profile. The direction of the MR jet was assessed from both parasternal and apical views, and the area of the largest clearly definable color flow disturbance was traced in each view as an index of the severity of MR (17,18). The width of the jet vena contracta as it emerges from the valve leaflets was measured in each view (19). Quantitative Doppler assessment of regurgitant volume, regurgitant fraction, and effective regurgitant orifice area (EROA) was performed using established methods (20). Similarly, the proximal isovelocity surface area method was also used to determine peak regurgitant flow rate and EROA (20). In patients without MR, vena contracta width and EROA were considered to be 0 for purposes of statistical analysis. Finally, pulsed Doppler spectra from the pulmonary veins were assessed in a modified four-chamber view for systolic flow reversal (20). All measurements of MR severity were made at the core laboratory; results were sent to the data coordinating center where all statistical analysis was performed.
Statistical analysis.
Continuous variables are reported as mean ± one standard deviation. Group comparisons were made by paired t test or Wilcoxon rank sum test, as appropriate. Categorical variables are reported as proportions; comparisons were made by chi-square test or Fisher exact test, as appropriate. The primary end point was death; the secondary end point was death, transplant, or hospitalization for HF. Survival was determined by the Kaplan-Meier method and p values were calculated by the log-rank test. All echocardiographic variables were assessed for univariate statistical significance using a Cox proportional hazard regression model. Variables with a p value <0.1 were then entered into a multivariable model, which was adjusted for the baseline covariates age, diabetes, creatinine, NYHA functional class, and treatment group. Optimal cut points for variables that were predictive of outcome on multivariable analysis were selected by examining box plots showing whether or not an event occurred. For all analyses, a p value
0.05 was considered statistically significant.
| Results |
|---|
|
|
|---|
|
|
120 ml/m2. Left ventricular end-diastolic volume index, mitral deceleration time, and vena contracta width (a marker of MR severity) were significant predictors of the combined end point of death, transplant, or hospitalization for CHF. Figure 2 shows the Kaplan-Meier event-free survival plots for these three variables using cut points of
120 ml/m2 for LV end-diastolic volume index,
150 ms for mitral deceleration time, and
0.4 cm for MR vena contracta width.
|
|
|
|
| Discussion |
|---|
|
|
|---|
35%). LV end-diastolic volume index. Left ventricular volumes are important predictors of outcome after acute myocardial infarction. White et al. (22) showed that LV end-systolic volume index was an independent predictor of survival after acute myocardial infarction. In the Survival And Ventricular Enlargement (SAVE) trial, LV cavity area, measured in a short-axis view, was reduced by captopril, and this reduction was associated with improved outcomes (23). Moreover, baseline LV cavity area predicted cardiovascular outcome regardless of treatment assigned (24). The SAVE trial also showed that LV remodeling after myocardial infarction was associated with ventricular arrhythmias (25).
Echocardiographic LV volume data has also been reported in HF clinical trials. In the SOLVD prevention and treatment arms (26), enalapril significantly reduced LV end-diastolic volume, LV end-systolic volume, and LV mass over 12 months compared with placebo in 301 patients enrolled in an echocardiographic substudy. Carvedilol has also been shown to reduce LV volumes and increase LVEF, in 123 patients in the Australia-New Zealand Heart Failure Research Collaborative Group (27). The Val-HeFT echocardiographic substudy (6) showed a reduction in LV end-diastolic dimension and an increase in LVEF with valsartan therapy compared to placebo in 5,010 patients. Although these studies showed favorable effects on LV remodeling of drugs known to improve survival in HF, they did not report the relation of echocardiographic variables measured at baseline to outcomes in HF, nor did they evaluate other important echocardiographic variables such as LV filling velocities or MR severity. In this substudy of the BEST trial, LV end-diastolic volume index was the only echocardiographic variable that predicted the primary outcome of death when adjusted for clinical covariates.
Many echocardiography laboratories do not measure LV volumes, but instead rely on LV dimensions, in part because of their ease of measurement and many years of clinical familiarity. In this study, LV dimensions were predictive of death and the combined end point on univariate, but not on multivariable, analysis. Recent studies using contrast echocardiography have shown that the accuracy of LV volumes is improved in technically difficult patients by the use of contrast (28,29), thus enabling measurement of LV volumes and LVEF in nearly all patients. The present data support routine measurement and reporting of LV volumes by echocardiography in patients with HF. Moreover, the prognostic importance of LV end-diastolic volume index highlights the interest in newer surgical techniques to reduce LV volumes in ischemic cardiomyopathy (30). A large clinical trial has been initiated to determine whether reducing LV volumes by surgical ventricular restoration will improve mortality in HF (31).
Mitral deceleration time. This study confirms the results of previous single-center studies that showed that mitral deceleration time predicts outcome in HF (8,9,32,33). In this large, multicenter trial of patients with advanced HF, mitral deceleration time independently predicted the combined end point of death, heart failure hospitalization, or transplant. The optimal cut point for mitral deceleration time was 150 ms. The pathophysiologic basis for this finding is that mitral deceleration time is a marker of increased LV stiffness (34) and decreased myocardial viability (35), factors associated with extensive fibrosis and necrosis.
MR vena contracta width.
Functional MR is a common complication of ischemic heart disease and is widely considered to contribute to symptoms and mortality. Two large clinical trials have shown that functional MR occurring either early or late (>16 days) after acute myocardial infarction is associated with increased mortality (36,37). However, until recently, studies in dilated cardiomyopathy were small and limited mainly to nonischemic etiology (10,3840). A recent retrospective study by Trichon et al. (41) reported that the angiographic severity of MR predicts mortality in patients with HF and LVEF
40%. This study confirms the prognostic importance of MR in advanced HF in a prospective clinical trial. Although several echocardiographic measures of MR severity predicted the combined end point on univariate analysis, only vena contracta width remained predictive on multivariable analysis, even when adjusted for clinical covariates. Patients with a vena contracta width >0.4 cm had a greater likelihood of death, transplant, or HF hospitalization. Importantly, a vena contracta of 0.4 cm is generally considered to be only moderate MR (19).
Study limitations. The patients in this substudy were not randomly selected from all eligible patients in the BEST study. Instead, they were enrolled from expert sites, which underwent specific training in the acquisition of this study protocol. Although this method assured high-quality data, we cannot be certain that these results apply to the overall BEST study population. In fact, the finding that treatment with bucindolol predicted improved survival in this substudy indicates that these patients are not representative of the overall BEST study population, because bucindolol was not associated with a survival benefit in the main trial. Moreover, because this study was initiated in 1996, tissue Doppler, strain-rate imaging, harmonic imaging, and contrast agents were not used. Finally, LVEF was predictive of outcome on univariable, but not on multivariable, analysis. This differs from prior studies (47). Likely reasons include the narrow and limited range of LVEF in this substudy, a relatively small number of patients, and the fact the LV end-diastolic volume, which was a statistically significant predictor, is a mathematical determinant of LVEF.
Conclusions. Echocardiographic predictors of outcome in advanced HF include LV end-diastolic volume index, mitral deceleration time, and vena contracta width. These variables indicate that LV remodeling, increased LV stiffness, and MR severity are independent predictors of outcome in patients with advanced HF.
| Appendix |
|---|
|
|
|---|
| Supplementary data |
|---|
|
|
|---|
| Footnotes |
|---|
| References |
|---|
|
|
|---|
This article has been cited by other articles:
![]() |
B G. Svealv, E L Olofsson, and B Andersson Ventricular long-axis function is of major importance for long-term survival in patients with heart failure Heart, March 1, 2008; 94(3): 284 - 289. [Abstract] [Full Text] [PDF] |
||||
![]() |
M. Enriquez-Sarano, V. T. Nkomo, and H. Michelena Principles and Practice of Echocardiography in Cardiac Surgery Card. Surg. Adult, January 1, 2008; 3(2008): 315 - 348. [Full Text] |
||||
![]() |
A. M. Dart, J. E. Otterstad, B.-A. Kirwan, J. D. Parker, S. de Brouwer, P. A. Poole-Wilson, and J. Lubsen Predictive value of local and core laboratory echocardiographic assessment of cardiac function in patients with chronic stable angina: The ACTION study Eur J Echocardiogr, August 1, 2007; 8(4): 275 - 283. [Abstract] [Full Text] [PDF] |
||||
![]() |
J. N. Kirkpatrick, M. A. Vannan, J. Narula, and R. M. Lang Echocardiography in Heart Failure: Applications, Utility, and New Horizons J. Am. Coll. Cardiol., July 31, 2007; 50(5): 381 - 396. [Abstract] [Full Text] [PDF] |
||||
![]() |
J.K. Oh Echocardiography in heart failure: Beyond diagnosis Eur J Echocardiogr, January 1, 2007; 8(1): 4 - 14. [Full Text] [PDF] |
||||
![]() |
T. G. Neilan, D. S. Jassal, T. M. Perez-Sanz, M. J. Raher, A. D. Pradhan, E. S. Buys, F. Ichinose, D. B. Bayne, E. F. Halpern, A. E. Weyman, et al. Tissue Doppler imaging predicts left ventricular dysfunction and mortality in a murine model of cardiac injury Eur. Heart J., August 1, 2006; 27(15): 1868 - 1875. [Abstract] [Full Text] [PDF] |
||||
![]() |
A. A. Chen, M. J. Wood, D. G. Krauser, A. L. Baggish, R. Tung, S. Anwaruddin, M. H. Picard, and J. L. Januzzi NT-proBNP levels, echocardiographic findings, and outcomes in breathless patients: results from the ProBNP Investigation of Dyspnoea in the Emergency Department (PRIDE) echocardiographic substudy Eur. Heart J., April 1, 2006; 27(7): 839 - 845. [Abstract] [Full Text] [PDF] |
||||
![]() |
A. N. DeMaria, O. Ben-Yehuda, D. Berman, G. K. Feld, G. S. Ginsburg, B. H. Greenberg, W. Y.W. Lew, D. Sahn, and S. Tsimikas Highlights of the Year in JACC 2005 J. Am. Coll. Cardiol., January 3, 2006; 47(1): 184 - 202. [Full Text] [PDF] |
||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | SUBSCRIPTIONS | CURRENT ISSUE | PAST ISSUES | CARDIOSOURCE | SEARCH | HELP | FEEDBACK |