American Heart Journal
Volume 148, Issue 3 , Page 530, September 2004

Lack of relationship between doppler indices of diastolic function and left ventricular pressure transients in patients with definite diastolic heart failure☆

  • Gerard P. Aurigemma, MD

      Affiliations

    • Department of Medicine, Division of Cardiology, University of Massachusetts Medical School, Worcester, Mass, USA
    • Corresponding Author InformationReprint requests: Gerard P. Aurigemma, MD, Room S3-860, University of Massachusetts Medical School, 55 Lake Avenue North, Worcester, MA, 01655, USA.
    • ,
  • Michael R. Zile, MD

      Affiliations

    • Department of Medicine, Division of Cardiology, Medical University of South Carolina, Charleston, SC, USA
    • ,
  • William H. Gaasch, MD

      Affiliations

    • Department of Medicine, Division of Cardiology, Lahey Clinic, Burlington, Mass, USA

Received 16 December 2002; accepted 11 January 2004.

Article Outline

Abstract 

Objectives

The purpose of this study was to compare invasive with noninvasive indices of diastolic function in a well-defined group of patients with diastolic dysfunction and a history of diastolic heart failure.

Background

Patients with heart failure and a normal left ventricular (LV) ejection fraction comprise a very large portion of the heart failure population and most are thought to have diastolic heart failure. While clinical and Doppler criteria for diastolic dysfunction and diastolic heart failure have been developed, there remains some controversy about the need for invasive cardiac catheterization and/or echo-Doppler evaluation of LV diastolic function. To date, there is no consensus as to the utility of these 2 methods in the diagnosis of diastolic heart failure.

Methods

Forty-seven patients (mean age 58 ± 11 years) with a history of congestive heart failure and preserved ejection fraction (≥50%) by echocardiography underwent a combined hemodynamic/echo-Doppler study. Patients with coronary disease were excluded. Invasive parameters of LV diastolic function (tau, LV diastolic pressures) and Doppler parameters (peak E, peak A, E/A ratio, isovolumic relaxation time, and E deceleration time) were measured using standard techniques.

Results

There was a close correlation between invasively-determined parameters (tau vs end diastolic pressure: r = 0.62, P < .001). The relationships between standard Doppler parameters and LV diastolic pressures were uniformly poor. However, the relationship between Doppler isovolumic relaxation time and tau improved considerably when patients were subgrouped by hemodynamic load.

Conclusions

Standard echo-Doppler indices of diastolic function correlate poorly with LV diastolic pressure transients. The diagnosis of diastolic heart failure cannot be made on the basis of a single echo-Doppler parameter but, rather, all parameters must be examined in concert and used in combination with clinical observations.

 

Patients with heart failure and a normal left ventricular (LV) ejection fraction comprise a very large portion of the heart failure population.1, 2, 3 Many, if not most, such patients are said to have diastolic heart failure. Diagnostic criteria for diastolic heart failure have been developed by a European study group,4 and by Vasan and Levy,5 and Doppler echocardiographic criteria have been published by others.6 However, there remains some controversy about the need for invasive cardiac catheterization and/or echo-Doppler evaluation of LV diastolic function, and there is no consensus as to the utility of these 2 methods in the diagnosis of diastolic heart failure. Because invasive measures of diastolic function are impractical in most patients, echo-Doppler methods are generally favored in clinical investigation and practice. The purpose of this study was to compare results of invasive and noninvasive techniques in a well-defined group of patients with diastolic dysfunction and a history of diastolic heart failure.

Back to Article Outline

Methods 

Definitions 

In this article we will use the term diastolic dysfunction when we refer to abnormal indices of LV diastolic function. Thus, patients with evidence of abnormal LV chamber stiffness, or abnormal echo-Doppler filling parameters, or prolonged myocardial relaxation can be said to exhibit diastolic dysfunction. The term diastolic heart failure is used in patients with the signs and symptoms of heart failure and a normal LV ejection fraction; diastolic dysfunction is virtually always present in such patients.7

This was a prospective, hemodynamic and echocardiographic study of patients with clinically-defined diastolic heart failure. Some data from this study have been published previously.7 All measurements were made by blinded investigators in core laboratories.

Patient population 

The screening criteria included a history of heart failure and normal LV ejection fraction. Patients who met these criteria and who had been scheduled for diagnostic cardiac catheterization were then evaluated for participation in the study. Those who met the Framingham criteria for congestive heart failure (CHF) were potential candidates.8 A contemporaneous echocardiogram was then performed; enrollment required evidence of a normal LV chamber dimension (<55 mm), combined with LV wall thickness >11 mm, relative wall thickness >0.45, or LV mass >125 g/m2.9 Thus, stable patients with heart failure and echocardiographic evidence suggesting hypertrophic LV remodeling and an ejection fraction >50% were invited to participate in the study. Doppler measurements of LV filling were not utilized as screening criteria. The study was approved by the institutional review board of all participating hospitals.

Specific exclusion criteria included concurrent severe systemic disease, evidence of coronary heart disease (including LV asynergy by echocardiography or ventriculography, or a history of previous coronary bypass surgery), significant congenital or valvular disease, or known cardiomyopathy. We also excluded patients with recent hemodynamic instability; those who had been treated with dopamine, dobutamine, or another positive inotropic agent within 48 hours; and those with clinically significant atrial or ventricular arrhythmia, electronic pacemakers, or implantable cardiac defibrillators.

Forty-seven consecutive patients meeting the above criteria (27 men and 20 women, mean age 58 ± 11 years) comprised the study population; all 47 underwent combined echocardiographic-hemodynamic (micromanometer) studies.

Cardiac catheterization 

Cardiac catheterizations were performed using standard techniques. All patients were treated with benzodiazepines; other medications were withheld, and patients fasted for 12 hours before catheterization. A high-fidelity micromanometer pigtail catheter was placed into the LV under fluoroscopic guidance. Before insertion, the micromanometer catheter was precalibrated in warm saline. After insertion, calibration was confirmed, and the catheter was recalibrated if necessary. Doppler and LV echocardiographic recordings were then obtained simultaneously with the acquisition of LV pressure data. We measured LV systolic pressure, diastolic pressure, and the time constant of isovolumic pressure decline (tau). LV early diastolic pressure was defined as the lowest pressure after mitral valve opening; LV pre-A-wave pressure was defined as the LV pressure midway through diastole; LV end-diastolic pressure (LVEDP) was defined as the pressure after atrial contraction just before LV systolic pressure rise.4 LV pressure data were digitalized at 5-ms intervals, and the relaxation time constant was calculated with the method of Weiss et al.10

Echocardiography 

Echocardiographic data were obtained using standard 2.5- to 3.5-MHz transducers and standard equipment with settings adjusted to optimize visualization of the ventricular endocardial contours while avoiding excessive gain artifact. LV dimensions and wall thickness were measured according to the recommendations of the American Society of Echocardiography by use of the leading edge convention, and calculations were made with previously published methods.7, 11 Pulsed Doppler interrogation of mitral inflow was performed with the sample volume between the tips of the mitral leaflets in the apical 4-chamber view with the use of 1- to 2-mm sample volume aligned with color inflow. Images were recorded on super VHS tape and recording paper (100 mm/s) for measurement.

Data analysis 

Data are presented as mean ± SD. Relationships between 2 variables were tested by linear regression analysis. A P value of < .05 was considered statistically significant. Invasive and noninvasive parameters of diastolic filling were compared using linear regression analysis. Because Doppler mitral inflow parameters are known to be influenced by both the rate of relaxation as well as the level of LV filling pressure,12, 13, 14, 15, 16, 17 we attempted to account for these influences in separate analyses. Therefore, IVRT, peak E wave, and deceleration time were all correlated with tau, using linear regression analysis, and the study population was subdivided by LVEDP to help account for the influence of filling loads. All statistical operations were performed using SPSS for Windows 9.0 (SPSS, Chicago, Ill).

Back to Article Outline

Results 

All 47 patients had a history of CHF and a normal LV ejection fraction, and they met the Vasan and Levy criteria for “definite” diastolic heart failure.5 This was a population exhibiting concentric LV geometry with a relative wall thickness exceeding 0.45 and ≥1 abnormal indices of diastolic function.7

Relations between invasive indices 

As is shown in Figure 1, there was a strong positive correlation between LVEDP and tau (r = 0.62, P < .001). A similar positive correlation between LV early diastolic pressure and tau was seen (r = 0.66, P < .001). Thus, a slow LV relaxation rate (increased tau) was strongly related to elevated LV diastolic pressures. This, however, does not necessarily indicate cause and effect (see Discussion).

  • View full-size image.
  • Figure 1. 

    Relation between LV relaxation rate (tau) and end-diastolic pressure. There is a strong positive correlation between relaxation rate and filling pressures (r = 0.62, P < .001), but this does not necessarily imply a cause and effect. See text for details.

Relations between invasive and noninvasive indices 

There was no significant relation between E wave velocity or E/A ratio and the LVEDP (Figure 2). Likewise, there was no significant relation between the E wave deceleration time or the isovolumic relaxation time and the LVEDP (Figure 3). The results were similar when these echo-Doppler indices were plotted against LV early diastolic pressure. As is shown in Figure 4, the relation between IVRT and tau depends on the LVEDP. Thus, for any given value for tau, the IVRT is longer when the filling pressures are low (<20 mm Hg) and shorter when filling pressures are high (>30 mm Hg). The slope of the relation between IVRT and tau is significantly higher (P < .001) when the LVEDP is low than when the LVEDP is high. These relations illustrate the interactions between relaxation parameters and hemodynamic loads.

  • View full-size image.
  • Figure 2. 

    Relation between peak E wave velocity and end-diastolic pressure (A), and the E/A ratio and end-diastolic pressure (B). There is no significant relation between the Doppler parameters and end-diastolic pressure (both P = NS).

  • View full-size image.
  • Figure 3. 

    Best fit relation between the E wave deceleration time and end-diastolic pressure (A), and the isovolumic relaxation time and end-diastolic pressure (B). There is no significant relation between the Doppler parameters and end diastolic pressure (both P = NS).

  • View full-size image.
  • Figure 4. 

    Relation between isovolumic relaxation time and relaxation rate (tau) for three ranges of filling pressures (<20, 20–30 and >30 mm Hg). The slope of the time versus rate relation is higher when the filling pressure is low; the slope is lower when the filling pressure is elevated (A). In B, the slope of this relation is shown for the 3 subgroups. Thus, at any common value for tau, the isovolumic relaxation time is longer when the filling pressures are low and shorter when the filling pressures are high.

Back to Article Outline

Discussion 

The purpose of this study was to evaluate the relations between LV diastolic pressure transients, including the relaxation time constant, and the traditional echo-Doppler indices of LV diastolic function. We studied a group of patients with well-documented diastolic dysfunction, a history of heart failure, and a normal LV ejection fraction; all patients had been treated and were stable at the time of the study.7 In this population of patients with “definite diastolic heart failure”,5 the echo-Doppler indices of diastolic function correlate poorly with pressure-derived indices of LV diastolic function.

The reason for these poor correlations is likely related to the sensitivity of the measured parameters to hemodynamic loads. For example, the lack of a relation between IVRT and tau appears to be related to such loads (Figure 4). In this example, the relation between IVRT and tau at different LV filling pressures indicates that for any given value for tau, the IVRT was shorter when the filling pressures were >30 mm Hg and longer when the filling pressures were <20 mm Hg (Figure 4). A higher filling pressure is likely to be associated with a higher mitral valve opening pressure and, therefore, earlier opening of the valve. As a result, the IVRT shortens. By contrast, lower filling pressures and lower mitral valve opening pressures should result in later opening of the valve and a longer IVRT. The sensitivity of IVRT to acute and short-term hemodynamic interventions is well recognized, but our results confirm a serious limitation to the use of IVRT as a relaxation index in treated and hemodynamically stable patients with definite diastolic dysfunction.

Because of the high prevalence of abnormal relaxation rates (increased tau) and the strong positive correlation between tau and LVEDP (Figure 1), we considered the possibility that incomplete relaxation caused or contributed to the high prevalence of elevated LVEDP. Because tau is equal to the time required for pressure to fall from its initial level to 37% of that level, it follows that pressure would fall to <3% of the initial level in <3.5 tau, relaxation would be incomplete at end diastole and the LVEDP could be affected. For example, if tau were 50 to 60 ms, the LVEDP could be affected only if the diastolic interval were <175–210 ms—a period seen only during tachycardia, which was not present in any of our patients. We therefore conclude that the elevated LVEDP seen in our patients was not a consequence of the abnormally prolonged myocardial relaxation. Rather, the high LVEDP was most likely caused by increased relative wall thickness and abnormal passive stiffness of the LV chamber.

Our results differ somewhat from some prior published clinical work comparing invasive and noninvasive indices of diastolic function. One reason for these differences may be the presence or absence of coronary heart disease in the study population; we excluded those patients from our study. Stoddard14 compared Doppler mitral inflow data with micromanometer pressure data in patients with and without coronary disease. In the subgroup of patients without significant coronary disease, peak E and the E/A ratio correlated with tau (r = 0.51 and r = 0.28, respectively). The correlation coefficients were higher in those patients with coronary artery disease and incomplete relaxation (tau = 84 ± 14). Appleton15 studied a consecutive series of patients undergoing catheterization with both Doppler and invasive hemodynamics; most patients had some degree of coronary disease, and the mean ejection fraction was in the normal range. They showed a good correlation between peak E and LVEDP (r = 0.50) and an even closer relationship between the Doppler E/A ratio and LVEDP (r = 0.73). Similarly, Mulvagh et al studied a series of patients with a wide range of ejection fractions (mean 55 ± 15, range 15%–75%) and showed a good correlation between E/A ratio and IVRT and LVEDP (r = −0.53, −0.73, respectively).16 However, such good correlations have not always been found. Oki et al17 studied 50 individuals, using high fidelity micromanometer-tip catheters and Doppler mitral inflow analyses. There was virtually no discernible relationship between the peak E or the E wave deceleration time and tau in the subpopulation with elevated diastolic pressures,17 a finding remarkably similar to the data presented herein.

The poor correlation between peak E and LV filling pressures in a population of patients with normal EF is perhaps explained by considering that the E wave height correlates with the transmitral pressure gradient in early diastole.12 In a population restricted only to patients with preserved EF, it is likely that some individuals will have a high E velocity due more to a small end-systolic volume and vigorous diastolic suction (low early diastolic pressures) than to high left atrial pressure. Therefore, the correlation between filling pressure and E wave in a group with normal EF might be expected to be poorer in such a population than in studies with a sizeable proportion of patients with significant LV systolic dysfunction, in whom a high E wave velocity is undoubtedly due to high left atrial pressure.

It should be emphasized that our invasive and noninvasive data were measured by blinded investigators, a method not used in other published studies. Clearly, the field of noninvasive estimation of LV filling pressures is evolving, with newer techniques applied to this study. Therefore, we do not imply that the poor correlations between invasive and noninvasive measures seen with standard Doppler measures will apply to results obtained with, for example, tissue Doppler imaging.

In this study, we found that echo-Doppler indices of diastolic function correlate poorly with LV diastolic pressure transients. These findings are likely due (at least in part) to the impact of different hemodynamic loads. We conclude that the diagnosis of diastolic heart failure cannot be made on the basis of a single echo-Doppler parameter but, rather, parameters must be examined in concert, and used in combination with clinical observations.

Back to Article Outline

Acknowledgements 

We express our appreciation to Jackie Jolie for assistance in manuscript preparation.

Back to Article Outline

Appendix. 

Study site, principal investigator, associate investigators, and nurse coordinators were: Medical University of South Carolina and the Ralph H. Johnson Veterans Affairs Medical Center: Michael R. Zile, MD, Christopher D. Nelson, MD, Melia Knotts, RN, Joan Zile, RN, and Leslie Harrell, RN; Lahey Clinic Medical Center: William H. Gaasch, MD, and Robin Sgrosso; University of Colorado Health Sciences Center: John D. Carroll, MD, JoAnn Lindenfeld, MD, Kathy Kioussopoulos, RN, and Keith Hellman; University of Texas Health Science Center San Antonio: Marc D. Feldman, MD, John Erikson, MD, PhD, Teresa Huber, RN, and Mary Alice Garcia, RDCS; University of Massachusetts Medical Center: Gerard P. Aurigemma, MD, Theo E. Meyer, MD, PhD, Eugene S. Chung, MD, and Kathy Coleman, RN; Rush Medical College, Rush-Presbyterian-St Luke's Medical Center: Joseph Parrillo, MD, Gary L. Schaer, MD, R. Jeffrey Snell, MD, Clifford Kavinsky, MD, Carolyn Ault, RN, Tony Hursey, MPH, Philip R. Liebson, MD, and Joanne Sandelski, RDMS; Cardiac Centers of Louisiana, LLC: Jalal Ghali, MD, Tommy Brown, MD, James Smith, MD, and Lela Parks, RN; and Mitsubishi Chemical America: David Katz, PhD, and Connie Colonnese, RN.

Back to Article Outline

References 

  1. Gaasch WH. Diagnosis and treatment of heart failure based on left ventricular systolic or diastolic dysfunction. JAMA. 1994;271:1276–1280
  2. Vasan S, Larson MG, Benjamin EJ, et al.  Congestive heart failure in subjects with normal versus reduced ejection fraction (prevalence and mortality in a population based cohort). J Am Coll Cardiol. 1999;33:1948–1955
  3. Aurigemma GP, Gottdiener JS, Shemanski L, et al.  Predictive value of systolic and diastolic function for incident congestive heart failure in the elderly (The Cardiovascular Health Study). J Am Coll Cardiol. 2001;37:1042–1048
  4. Paulus WJ  European Study Group on Diastolic Heart Failure . How to diagnose diastolic heart failure. Eur Heart J. 1998;19:990–1003
  5. Vasan RS, Levy D. Defining diastolic heart failure (a call for standardized diagnostic criteria). Circulation. 2000;101:2118–2121
  6. Rakowski H, Appleton C, Chan KL, et al.  Canadian consensus recommendations for the measurement and reporting of diastolic dysfunction by echocardiography. J Am Soc Echocardiogr. 1996;9:736–760
  7. Zile MR, Gaasch WH, Carroll JD, et al.  Heart failure with a normal ejection fraction (is measurement of diastolic function necessary to make the diagnosis of diastolic heart failure?). Circulation. 2001;104:779–782
  8. McKee PA, Castelli WP, McNamara PM, et al.  The natural history of congestive heart failure, the Framingham study. N Engl J Med. 1971;285:1441–1446
  9. Ganau A, Devereux RB, Roman MJ, et al. Patterns of left ventricular hypertrophy and geometric remodeling in essential hypertension. J Am Coll Cardiol 1992;19:1550–8
  10. Weiss JL, Fredriksen JW, Weisfeldt ML. Hemodynamic determinants of the time course of fall in canine left ventricular pressure. J Clin Invest. 1976;58:751–756
  11. Aurigemma GP, Gaasch WH, Villegas B, et al.  Noninvasive assessment of left ventricular mass, chamber volume, and contractile function. Curr Probl Cardiol. 1995;20:361–440
  12. Nishimura RA, Tajik AJ. Evaluation of diastolic filling of left ventricle in health and disease (Doppler echocardiography is the clinician's rosetta stone). J Am Coll Cardiol. 1997;30:8–18
  13. Garcia M, Thomas JD, Klein AL. New Doppler echocardiographic applications for the study of diastolic function. J Am Coll Cardiol. 1998;32:865–875
  14. Stoddard MF, Pearson AC, Kern MJ, et al.  Left venticular diastolic function (comparison of pulsed Doppler echocardiographic and hemodynamic indexes in subjects with and without coronary artery disease). J Am Coll Cardiol. 1989;13:327–336
  15. Appleton CP, Galloway JM, Gonzalez MS, et al.  Estimation of left ventricular filling pressures using two-dimensional and Doppler echocardiography in adult patients with cardiac disease. Additional value of analyzing left atrial size, left atrial ejection fraction and the difference in duration of pulmonary venous and mitral flow velocity at atrial contraction. J Am Coll Cardiol. 1993;22:1972–1982
  16. Mulvagh S, Quinones MA, Kleiman NS, et al.  Estimation of left ventricular end-diastolic pressure from Doppler transmitral flow velocity in cardiac patients independent of systolic performance. J Am Coll Cardiol. 1992;20:112–119
  17. Oki T, Tabata T, Yamada H, et al.  Clinical application of pulsed Doppler tissue imaging for assessing abnormal left ventricular relaxation. Am J Cardiol. 1998;81:663

☆ Supported by a grant from Mitsubishi Chemical America, Inc, White Plains, NY.

PII: S0002-8703(04)00132-2

doi:10.1016/j.ahj.2004.01.022

American Heart Journal
Volume 148, Issue 3 , Page 530, September 2004

Article Tools

* Image Usage & Resolution