Mitral Annulus Velocity in the Estimation of

Left Ventricular Filling Pressure:

Prospective Study in 200 Patients
Yong-Jin Kim, MD, and Dae-Won Sohn, MD, Seoul, Korea

The ratio of early mitral inflow to early mitral annular was not dependent on the left ventricular systolic
velocity (E/E´) was suggested as a useful index in the function (ejection fraction [EF] ≥50%: r = 0.74, P < .001
estimation of left ventricular filling pressure. versus EF <50%: r = 0.70, P < .001). The E/E´ ratio of ≥9
This study was performed to validate the clinical best discriminated elevated (>12 mm Hg) from normal
usefulness of E/E´ ratio in a large number of patients. left ventricular pre-A pressure with a sensitivity of
Simultaneous left ventricular pressure measurements 81% and a specificity of 80%. Our study results also
and Doppler examinations were performed in 200 showed that quantitative estimation of left ventricular
consecutive patients at the cardiac catheterization pre-A pressure could be suggested by the simplified
laboratory. The E/E´ ratio correlated well with pre–A- equation of pre-A pressure = E/E´ + 4 with reasonable
wave pressure (r = 0.74, P < .001), and the correlation accuracy. (J Am Soc Echocardiogr 2000;13:980-5.)

Estimation of left ventricular filling pressure is an new index also depends on left ventricular systolic
important part of the evaluation of patients with function for its diagnostic accuracy.
symptoms of congestive heart failure. Mitral inflow
parameters have been most widely used for this
purpose with well-known limitations in patients METHODS
with normal left ventricular systolic function.1-5
Pulmonary venous flow patterns may provide addi-
Study Subjects
tional information, but these flow patterns cannot be
obtained in all patients. Recently, Garcia et al6 showed Simultaneous left ventricular pressure measurements and
that left ventricular filling pressure could be estimat- Doppler examinations were performed in 200 consecutive
ed by the ratio of early mitral inflow velocity to color patients who were undergoing clinically indicated left ven-
M-mode propagation velocity. They intended to cor- triculography or coronary angiography. Patients with
rect early mitral inflow velocity for the influence of arrhythmia, valvular stenosis, significant valvular regurgita-
left ventricular relaxation with propagation velocity. tion, regional wall motion abnormality at the basal septum,
Early diastolic mitral annulus velocity (E´) has been or apical dyskinesis were excluded.The mean age was 61 ±
reported to be a relatively load-independent parame- 9 (range 31 to 84) years, and 127 (64%) patients were men.
ter reflecting left ventricular relaxation.7-11 Therefore, Clinical indications for the invasive studies were angina in
E´ could be substituted for the propagation velocity, 96 patients,myocardial infarction in 39,follow-up study after
and left ventricular filling pressure could be estimat- coronary artery bypass surgery or coronary intervention in
ed by the E/E´ ratio.12 This study was performed to 38, atypical chest pain in 17, and other conditions in 10.
validate the clinical usefulness of the E/E´ ratio in a
large number of patients and to test whether this Echocardiography
Echocardiograms were obtained with an Acuson XP/10
From the Clinical Research Institute and Division of Cardiology, (Mountain View, Calif) echocardiographic system with a 2.5-
Department of Internal Medicine, Seoul National University MHz transducer. Two-dimensional echocardiography was
College of Medicine. performed, followed by a Doppler study. The sample volume
Reprint requests: Dae-Won Sohn, MD, Division of Cardiology, (2.5 mm) of the pulsed wave Doppler was placed between
Department of Internal Medicine, Seoul National University
College of Medicine, 28 Yongun-Dong, Chongno-Gu, Seoul the tips of the mitral leaflets on the apical 4-chamber view.
110-744, Korea. Peak early (E) and late (A) mitral inflow velocities, E/A ratio,
Copyright © 2000 by the American Society of Echocardiography. and deceleration time of E velocity (DT) were obtained.
0894-7317/2000/$12.00 + 0 27/1/107156 Pulsed wave tissue Doppler imaging (TDI) was performed
doi:10.1067/mje.2000.107156 by activating the TDI function in the same machine. A sam-

980
Journal of the American Society of Echocardiography
Volume 13 Number 11 Kim and Sohn 981

Figure 1 Examples of measuring E/E´ ratio. Patients

with normal (left panels) and elevated (right panels) left
ventricular filling pressure. E, Early mitral inflow velocity;
A, late mitral inflow velocity; E´, early mitral annulus
velocity; A´, late mitral annulus velocity.

ple volume of the same size was located at the septal side of B
the mitral annulus. Peak early (E´) and late (A´) diastolic
mitral annular velocities and the E´/A´ ratio were obtained.
Doppler echocardiograms were recorded on a strip
chart with the sweep speed of 100 mm/s. The E/E´ ratio
(Figure 1) was obtained with the mean values E and E´ in
5 consecutive cardiac cycles.

Cardiac Catheterization
Left-heart catheterization was performed through the
femoral approach. The study was performed before the
left ventriculography or coronary angiography. A 7F pig-
tail catheter was introduced into the left ventricle. Care
was taken to avoid premature ventricular contraction.
Left ventricular diastolic pressures before atrial contrac-
tion (pre-A) were measured. To exclude the respiratory
variation, pressures were measured during end-expirato- C
ry apnea. Mean values of the 5 consecutive cardiac cycles Figure 2 Correlation between deceleration time of early
were used in the analysis. mitral inflow (DT) and pre-A pressure. A, Correlation in the
entire group of patients. B, Correlation in patients with nor-
Statistics mal (ejection fraction ≥50%) left ventricular systolic function.
C, Correlation in patients with depressed left ventricular sys-
Continuous variables are presented as mean ± SD.Sensitivity tolic function. LVEF, Left ventricular ejection fraction; pre-
and specificity were calculated with the standard formula. A, left ventricular diastolic pressures before atrial contraction.
 Journal of the American Society of Echocardiography
982 Kim and Sohn November 2000

Linear regression analysis was used to correlate Doppler

parameters to the left ventricular pre-A pressure, and the
95% confidence interval (CI) of the correlation coefficient
was calculated with the Fisher method. All statistical analy-
sis was performed with a SAS (Cary, NC) statistical package.
A P value less than .05 was considered to be significant.

RESULTS

Baseline Characteristics

Adequate mitral inflow parameters and mitral annu-

lus velocity could be obtained in all patients. Left
A ventricular systolic function was normal (ejection
fraction [EF] ≥50%) in 167 patients (84%) and
depressed (EF <50%) in 33 patients, with a mean
ejection fraction of 57% ± 10%. Mean pre-A pressure
was 12.4 ± 4.6 mm Hg, and 80 patients (40%) had a
pre-A pressure of ≥12 mm Hg.
Mitral Inflow Pattern and Left Ventricular
Filling Pressure
The DT only weakly correlated with left ventricular
pre-A pressure (r = –0.39, P < .01). Correlation
between DT and left ventricular pre-A pressure was
dependent on the left ventricular systolic function: a
higher r value was found in patients with a depressed
left ventricular systolic function (r = –0.67; 95%CI:
–0.59 to 0.74) than in patients with a normal left ven-
B tricular systolic function (r = –0.21; 95%CI: –0.07 to
0.34) (Figure 2). Pre-A pressures were >12 mm Hg in
all patients (N = 12) with restrictive mitral inflow pat-
tern (DT ≤ 150 ms, E/A ratio ≥2). Thus a restrictive
mitral inflow pattern was highly specific (specificity
100%) but not sensitive (sensitivity 15%) in predicting
elevated left ventricular filling pressure. A relaxation
abnormality pattern in the mitral inflow has been usu-
ally regarded as representing low or normal left ven-
tricular filling pressure. However, among the 54
patients with relaxation abnormality in the mitral
inflow pattern (DT ≥240 ms, E/A ratio <1), 18 patients
(33%) showed a pre-A pressure of ≥12 mm Hg.
E/E´ Ratio in the Estimation of Left Ventricular
Filling Pressure
C Among the Doppler parameters measured in our study,
Figure 3 Correlation between E/E´ ratio and pre-A pres- the E/E´ ratio correlated best with pre–A-wave pressure
sure. A, Correlation in the entire group of patients. B, (r = 0.74, P < .001). Correlation was not significantly
Correlation in patients with normal (ejection fraction affected by the left ventricular systolic function (Figure
≥50%) left ventricular systolic function. C, Correlation in 3) or mitral inflow pattern (Figure 4). In predicting the
patients with depressed left ventricular systolic function.
Pre-A, Left ventricular diastolic pressures before atrial con-
pre–A-wave pressure, the mean difference between
traction; E, early mitral inflow velocity; E´, early mitral Doppler-predicted and catheter-measured pressures
annulus velocity; LVEF, left ventricular ejection fraction. was 0.06 ± 3.1 mm Hg (Figure 5). In only 18 patients
Journal of the American Society of Echocardiography
Volume 13 Number 11 Kim and Sohn 983

Figure 4 Correlation between E/E´ ratio and pre-A pres-

sure in patients with relaxation abnormality in the mitral
inflow pattern. Pre-A, Left ventricular diastolic pressures
before atrial contraction; E, early mitral inflow velocity; E´,
early mitral annulus velocity.

(9%), the difference was more than 5 mm Hg.An E/E´

ratio of ≥9 could best discriminate elevated (>12 mm
Hg) from normal pre-A pressure with a sensitivity of
81% and a specificity of 80%.Applying cutoff values of B
≥8 and ≥10 showed sensitivities of 95% and 52%, and
Figure 5 A, Correlation between Doppler-predicted and
specificities of 59% and 91%, respectively.
catheter-measured left ventricular filling pressures. B,
Difference between Doppler-predicted and catheter-mea-
sured left ventricular filling pressures plotted against the
DISCUSSION average of both observations. The middle line indicates
the average difference between the two methods, and the
outer lines represent 95% limits of agreement. Pre-A, Left
Our study confirmed the results of previous studies
ventricular diastolic pressures before atrial contraction.
that showed that the E/E´ ratio is a reasonably good
index for predicting elevated left ventricular filling
pressure.12-14 Moreover, this new index has several left ventricular filling pressure.18 The difference
advantages over conventional Doppler parameters. between mitral and pulmonary A-wave durations is
First, this index could be obtained in all patients useful in the estimation of left ventricular filling pres-
included in our study. In the previous reports, ade- sure,15,19,20 and this parameter is reported to be
quate pulmonary venous flow signals, for example, independent of left ventricular systolic function.
could not be obtained in up to 18% of subjects with- However, this index could only give information
out cardiac disease15 and in up to 84% of patients in about left ventricular end-diastolic pressure. Elevated
intensive care units.16 We think the most important end-diastolic pressure is not always associated with
prerequisite of any diagnostic parameter is that it elevated mean left atrial pressure, which is more
can be obtained in all patients. Second, our study closely represented by the pre-A pressure.20,21 Third,
results showed that the E/E´ ratio is useful in pre- conventional Doppler parameters could only give
dicting elevated left ventricular pre-A pressure, qualitative estimation of elevated left ventricular fill-
regardless of left ventricular systolic function or ing pressure. Complex equations that use conven-
mitral inflow pattern. Conventional mitral inflow tional Doppler parameters were proposed for the
parameters are most helpful only in patients with a quantitative estimation of left ventricular filling pres-
depressed left ventricular systolic function.17 sure.22,23 However, those equations could not be
Moreover, when patients show a relaxation abnor- used easily in routine clinical practice. With the
mality in the mitral inflow pattern, mitral inflow pat- regression equation pre-A pressure = (0.95 E/E´) +
tern alone cannot give any information about the 3.8, pre-A pressures were more than 5 mm Hg apart
 Journal of the American Society of Echocardiography
984 Kim and Sohn November 2000

from the predicted value in only 18 patients (9%) in motion abnormalities of the basal septum because
our study. Because the usual range of variation in the mitral annulus velocity could theoretically be influ-
E/E´ ratio is from 5 to 30, this equation could be fur- enced by these regional wall motion abnormalities.
ther simplified to pre-A pressure = E/E´ + 4. When Second, we included the selected group of patients
this simplified equation was applied, predicted values who were candidates for invasive study.This is the lim-
were within 3 mm Hg of the measured values in 153 itation intrinsic to all studies that use invasive hemo-
patients (77%) and differed more than 5 mm Hg in dynamics as a gold standard.Third, pre-A pressure was
only 16 patients (8%). This simplified equation pro- obtained with a fluid-filled catheter. Because of the
vides the quantitative estimation of left ventricular fill- potential for overdampening and underdampening
ing pressure, which can be easily applied in clinical in fluid-filled catheters, a micromanometer-tipped
practice. Fourth, the E/E´ ratio is useful where con- catheter would be ideal. However, this method is the
ventional Doppler parameters cannot be applied.The standard in measuring pressures in most clinical set-
E/E´ ratio could still be used when mitral E and A tings. Fourth, in many clinical situations, such as
waves are fused because of tachycardia.13,24 patients in intensive care units, monitoring the serial
changes of left ventricular filling pressure is important.
Previous Studies Whether E/E´ can reflect acute changes in left ventric-
Nagueh et al demonstrated good correlation between ular filling pressure should be studied in the future.
the E/E´ ratio and pulmonary capillary wedge pres-
sure (PCWP) in 60 patients with a mitral inflow pat-
tern of relaxation abnormality or pseudonormaliza- REFERENCES
tion (r = 0.87, P < .001)12 and also in 120 patients
1. Oh JK, Appleton CP, Hatle LK, Nishimura RA, Seward JB,
with sinus tachycardia in whom mitral E and A waves Tajik AJ. The noninvasive assessment of left ventricular dias-
were fused (r = 0.86, P < .001).13 In both studies, the tolic function with two-dimensional and Doppler echocar-
value of >10 best predicted the PCWP of >12 mm Hg. diography. J Am Soc Echocardiogr 1997;10:246-70.
Sundereswaran et al14 also reported good correlation 2. Nishimura RA, Tajik AJ. Evaluation of diastolic filling of left
ventricle in health and disease: Doppler echocardiography is
in 50 patients with a heart transplant (r = 0.8,P < .01).
the clinician’s Rosetta stone. J Am Coll Cardiol 1997;30:8-18.
In their study, the value of ≥8 best predicted the 3. Appleton CP, Hatle LK, Popp RL. Demonstration of restric-
PCWP of ≥15 mm Hg. Our study showed a lower r tive ventricular physiology by Doppler echocardiography. J
value and a different cutoff value of E/E´ ratio (≥9) to Am Coll Cardiol 1988;11:757-68.
predict a pre-A pressure of ≥12 mm Hg.This discrep- 4. Vanoverschelde JJ, Robert AR, Gerbaux A, Michel X, Hanet
C, Wijin W. Noninvasive estimation of pulmonary arterial
ancy is likely the result of the following several differ-
wedge pressure with Doppler transmitral flow velocity pattern
ences. First, pre-A pressure was measured to represent in patients with known heart disease. Am J Cardiol 1995;75:
the left ventricular filling pressure, instead of PCWP, 383-9.
which was measured in the previous studies. Mean 5. Thomas JD, Choong CYP, Flachskampf FA, Weyman AE.
left atrial pressure usually has been estimated by Analysis of the early transmitral Doppler velocity curve: effect
of primary physiologic changes and compensatory preload
PCWP,but a significant difference could exist between
adjustment. J Am Coll Cardiol 1990;16:644-55.
two pressures,especially in the case of a high PCWP.25 6. Garcia MJ, Ares MA, Asher C, Rodriguez L, Vandervoort P,
Second, the sample volume for TDI was placed at the Thomas JD. An index of early left ventricular filling that com-
septal side of the mitral annulus in our study, whereas bined with pulsed Doppler peak E velocity may estimate cap-
it was at the lateral side of the mitral annulus in the illary wedge pressure. J Am Coll Cardiol 1997;29:448-54.
7. Assmann PE, Slager CJ, Dreysse ST, van der Borden SG,
previous studies. Nagueh et al12 observed a significant
Oomen JA, Roelandt JR. Two-dimensional echocardiograph-
relation between the early mitral annulus velocity ic analysis of the dynamic geometry of the left ventricle: the
measured at these two points (r = 0.88, P < .001) with basis for an improved model of wall motion. J Am Soc
slightly a higher value obtained at the lateral side. Echocardiogr 1988;1:393-405.
Third, study subjects were different. Previous studies 8. Rodriguez L, Garcia M, Ares M, Griffin BP, Nakatani S,
Thomas JD. Assessment of mitral annular dynamics during
included the patients who needed balloon-tipped pul-
diastole by Doppler tissue imaging: comparison with mitral
monary artery catheterization and probably included Doppler inflow in subjects without heart disease and in
more patients whose condition was hemodynamically patients with left ventricular hypertrophy. Am Heart J
unstable, which is reflected by the higher left ventric- 1996;131:982-7.
ular filling pressures. 9. Oki T, Tabata T, Yamata H, et al. Clinical application of
pulsed tissue Doppler imaging for assessing abnormal left
Study Limitations ventricular relaxation. Am J Cardiol 1997;79:921-8.
10. Sohn DW, Chai IH, Lee DJ, et al. Assessment of mitral annulus
There are several limitations in our study. First, we velocity by Doppler tissue imaging in the evaluation of left ven-
excluded the patients with apical dyskinesis and tricular diastolic function. J Am Coll Cardiol 1997;30:474-80.
Journal of the American Society of Echocardiography
Volume 13 Number 11 Kim and Sohn 985

11. Garcia MJ, Rodriguez L, Ares M, Griffin BP, Thomas JD. graphic evaluation of left ventricular filling pressures in
Differentiation of constrictive pericarditis from restrictive car- patients with cardiomyopathies: a simultaneous Doppler
diomyopathy: assessment of left ventricular diastolic velocities echocardiographic and cardiac catheterization study. J Am
in longitudinal axis by Doppler tissue imaging. J Am Coll Coll Cardiol 1996;28:1226-33.
Cardiol 1996;27:108-14. 19. Brunazzi MC, Chirillo F, Pasqualini M, et al. Estimation of
12. Nagueh SF, Middleton KJ, Kopelen HA, Zoghbi WA, left ventricular diastolic pressures from precordial pulsed-
Quinones MA. Doppler tissue imaging: a noninvasive tech- Doppler analysis of pulmonary venous and mitral flow. Am
nique for evaluation of left ventricular relaxation and estima- Heart J 1994;128:293-300.
tion of filling pressures. J Am Coll Cardiol 1997;30:1527-33. 20. Appleton CP, Galloway JM, Gonzalez MS, Gaballa M, Basnight
13. Nagueh SF, Kopelen HA, Middleton KJ, Quinones MA, MA. Estimation of left ventricular filling pressures using two-
Zoghbi WA. Doppler estimation of left ventricular filling dimensional and Doppler echocardiography in adult patients
pressure in sinus tachycardia. Circulation 1998;98:1644-50. with cardiac disease. J Am Coll Cardiol 1993;22:1972-82.
14. Sundereswaran L, Nagueh SF, Vardan S, et al. Estimation of left 21. Yamamoto K, Nishimura RA, Redfield MM. Assessment of mean
and right ventricular filling pressures after heart transplantation left atrial pressure from the left ventricular pressure tracing in
by tissue Doppler imaging. Am J Cardiol 1998;82:352-7. patients with cardiomyopathies. Am J Cardiol 1996;78:107-10.
15. Masuyama T, Lee JM, Tamai M, Tanouchi J, Kitabatake A, 22. Giannuzzi P, Imparato A, Temporelli PL, et al. Doppler-
Kamada T. Pulmonary venous flow pattern as assessed with derived mitral deceleration time of early filling as a strong
transthoracic pulsed Doppler echocardiography in subjects predictor of pulmonary capillary wedge pressure in postin-
without cardiac disease. Am J Cardiol 1991;18:65-71. farction patients with left ventricular systolic dysfunction. J
16. Nagueh SF, Kopelen HA, Zoghbi WA. Feasibility and accu- Am Coll Cardiol 1994;23:1630-7.
racy of Doppler echocardiographic estimation of pulmonary 23. Mulvagh S, Quinones MA, Kleiman NS, Cheirif J, Zoghbi WA.
artery occlusive pressure in the intensive care unit. Am J Estimation of left ventricular end-diastolic pressure from
Cardiol 1995;75:1256-62. Doppler transmitral flow velocity in cardiac patients indepen-
17. Yamototo K, Nishimura RA, Chaliki HP, Appleton CP, dent of systolic performance. J Am Coll Cardiol 1992;20:12-9.
Holmes DR, Redfield MM. Determination of left ventricular 24. Sohn DW, Kim YJ, Kim HC, Chun HG, Park YB, Choi YS.
filling pressure by Doppler echocardiography in patients with Evaluation of left ventricular diastolic function when mitral E
coronary artery disease: critical role of left ventricular systolic and A waves are completely fused: role of assessing mitral
function. J Am Coll Cardiol 1997;30:1819-26. annulus velocity. J Am Soc Echocardiogr 1999;12:203-8.
18. Nishimura RA, Appleton CP, Redfield MM, Ilstrup DM, 25. Walston A II, Kendall ME. Comparison of pulmonary wedge
Holmes DR, Tajik AJ. Noninvasive Doppler echocardio- and left atrial pressure in man. Am Heart J 1973;86:159-64.