(Circulation. 1997;96:2009-2015.)
© 1997 American Heart Association, Inc.
Articles |
From the Oregon Health Sciences University (M.I., T.S., S.R.H., D.J.S.), Portland; the National Heart, Lung, and Blood Institute (M.J., I.Y.), Bethesda, Md; and the Georgia Institute of Technology (R.S.H., A.P.Y.), Atlanta.
Correspondence to Michael Jones, MD, National Institutes of Health, Senior Surgeon and Investigator, National Heart, Lung, and Blood Institute, 9000 Rockville Pike, Bldg 14E Room 1074A, Bethesda, Md 20892.
| Abstract |
|---|
|
|
|---|
Methods and Results Twenty-nine hemodynamically different states were obtained pharmacologically in eight sheep with surgically induced aortic regurgitation. Instantaneous regurgitant flow rates (RFRs) were obtained with aortic and pulmonary electromagnetic flowmeters (EFMs), and aortic EROAs were determined from EFM RFRs divided by continuous wave Doppler velocities. Color Dopplerderived EROAs were estimated by measuring the maximal diameters of the CDVC. Peak and mean RFRs and regurgitant volumes per beat were calculated from vena contracta area continuous wave diastolic Doppler velocity curves. Peak EFM-derived RFRs varied from 1.8 to 13.6 (6.3±3.2) L/min (range [mean±SD]), mean RFRs varied from 0.7 to 4.9 (2.7±1.3) L/min, regurgitant volumes per beat varied from 7.0 to 48.0 (26.9±12.2) mL/beat, and the regurgitant fractions varied from 23% to 78% (55±16%). EROAs determined by using CDVC measurements correlated well with reference EROAs obtained by using the EFM method (r=.91, SEE=0.07 cm2). Excellent correlations and agreements between peak and mean RFR and regurgitant volumes per beat as determined by Doppler echocardiography and EFM were also demonstrated (r=.95 to .96).
Conclusions Our study indicates that the CDVC method can be used to quantify both aortic EROAs and regurgitant flow rates.
Key Words: echocardiography hemodynamics imaging valves
| Introduction |
|---|
|
|
|---|
| Methods |
|---|
|
|
|---|
Cardiac Catheterization and EFMs
The animals were instrumented for the
hemodynamic studies as follows. Bilateral transverse
thoracotomies were performed. A catheter was placed in the main
pulmonary artery via the femoral vein, and another catheter was
placed into the right common femoral artery for monitoring pressures
and blood gases. These catheters were interfaced with a
physiological recorder (ES 2000, Gould Inc) by
using fluid-filled pressure transducers (model PD23ID, Gould Statham).
Arterial blood gases and pH were maintained within
physiological ranges. Electromagnetic flow probes
(model EP455, Carolina Medical Electronics, Inc) were placed around the
skeletonized ascending aorta distal to the coronary ostia and
proximal to the brachiocephalic trunk and also on the main
pulmonary trunk above the pulmonary valve sinuses.
Calibration factors for the flow probes were corrected for the animals' hematocrits according to the manufacturer's specifications before each hemodynamic state. Occlusive zeros for the aortic and pulmonary probes were confirmed. Both aortic and pulmonary EFM records were displayed in the same multichannel recorder. To deal with zero baseline drift, the pulmonary artery flow zero baseline was adjusted according to the contour of its electromagnetic flow-probe signal; this baseline was reconfirmed by occlusive zeros. No animal had physiologically important pulmonary regurgitation. For determination of aortic and pulmonary flows, the integrals of aortic and pulmonary forward flows over time were determined by planimetry of the flow-signal recordings. Four consecutive cardiac cycles were analyzed for each hemodynamic measurement. The baselines for the aortic flow records were then adjusted until the forward minus the backward aortic flow volumes equaled the pulmonary forward flow volumes. The difference between the forward flows thus represented AR flow volume, as does the reversal phase of the aortic flow during diastole. Coronary arterial blood flow was measured in three sheep in a preliminary study and found to be small (0.13 to 0.23 L/min). As in other studies of AR, these values were considered negligible compared with the regurgitant volumes.9 RF was calculated as diastolic reverse aortic flow volume per minute divided by forward aortic flow volume per minute.
A hydrostatic standard was used for calibration of all pressure recordings. All hemodynamic recordings were performed simultaneously with the echocardiographic studies. After baseline measurements, varying degrees of severity of AR were produced by altering preload, afterload, or both by using blood transfusions and/or angiotensin II (Peptide Institute, Inc, provided by Tanabe Seiyaku Co). A total of 29 stable hemodynamic steady states (3 to 4 per animal) were obtained. Whole blood (usually 250 to 500 mL) was transfused to increase the pulmonary artery wedge pressure by 5 mm Hg. Alternatively, angiotensin II (2 µmol/mL) was infused at a rate necessary to increase the aortic diastolic pressure by 10 mm Hg. These strategies were implemented alone or in combination to produce and maintain an even new steady state.
Echocardiography
Echocardiography, including both color
Doppler flow mapping and CW spectral Doppler studies, was
performed with a Vingmed 775 system (Vingmed Sound, A/S) by using a
5-MHz, dynamically focused annular array transducer that used a 5-MHz
carrier frequency for imaging and 4 or 6 MHz for color and spectral
Doppler. The transducer was placed directly on the heart near the
apex. A pulse repetition frequency of 4.0 to 6.0 kHz was used for color
Doppler scanning. Gain settings were optimized for image quality by
using the maximal color gain level that would not introduce signals
outside areas of flow. Once established, depth and gain settings were
not changed during the recording period. Aliasing velocities of
0.44 to 0.94 m/s were selected for the initial imaging of both the AR
jet and the flow-convergence region. The aliasing velocity could be
changed by postprocessing software (EchoDisp, Vingmed Sound, A/S) after
the digital data were transferred to the microcomputer. Color sector
size was limited to 15° to 25° to allow frame rates up to 45
frames/s and to maximize angular line density for color Doppler
interrogation with a moderately high-quality setting (packet size).
When the image of the vena contracta was not discrete enough to
measure, we systematically reduced the frame rate to improve the
quality of the image. The area of interest was then magnified. The
color Doppler filter was held constant and set with a high-pass
filter to minimize velocities <8 to 16 cm/s. All color Doppler
echocardiographic images (as cine loops) and CW
Doppler traces (as time-scroll loops) were directly transferred in
digital format without analog conversion to a Macintosh II ci (Apple
Computer, Inc) for later digital analysis.
Color and CW Doppler Analysis
Color flow imaging of the vena contracta was performed in the
apical long-axis view. Considerable care was taken to measure the width
of the vena contracta as the smallest color flow diameter of the
narrowest portion of the high-velocity region just distal to the
orifice at the junction of the proximal flowconvergence acceleration
region and the variance-encoded turbulent regurgitant jet spray (Fig 1
). This transition was constantly and easily visualized
for flows coming toward the transducer consistently located in
the left ventricular outflow tract side of the aortic
valve, where the laminar aliasing flow-convergence zone meets the
mosaic turbulent proximal jet. This point was observed throughout the
cardiac cycle, and the largest cross-sectional area in
diastole was chosen. Individual determinations from three
cardiac cycles were measured and averaged. A circular configuration of
the EROA was assumed, and EROA was calculated by using the maximal vena
contracta width imaged in early diastole from the following
formula.
![]() |
|
CW Doppler traces of the AR jets were obtained from the apical as well as from other standard transducer positions under two-dimensional color Doppler echocardiographic guidance by using the transducer position most parallel to the AR flow that yielded the highest diastolic Doppler VTIs and regurgitant velocities. Doppler filters and gain settings were set to maximize and delineate the spectral Doppler envelope. The VTI of the AR flow was determined by planimetry of the area under the spectral Doppler velocity curve. All values were computed as an average of three consecutive beats.
RV/beat was calculated as the product of the EROA and the diastolic VTI. Thus, RV/beat was calculated as RV/beat=EROAxVTI of AR. Peak flow rate was calculated as peak flow rate=EROAxAR peak velocity and mean flow rate as mean flow rate=EROAxAR mean velocity.17 18
The EFM-derived reference peak EROA was also calculated by dividing the peak RFR by the corresponding CW Doppler velocity on the same beat, ie, reference peak EROA=(peak RFR by EFM)/(AR peak velocity by CW Doppler).
Interobserver and Intraobserver Variability
To evaluate the effect of observer variability on the EROA and
the RV/beat calculated from the CDVC, two independent observers (M.I.
and T.S.) analyzed 10 randomly selected
hemodynamic conditions at different times with the same
computer; each observer individually selected the frames to measure and
had no knowledge of the results obtained by the other observer or of
the electromagnetic flow data at the time. The EROA measurements were
repeated by one observer (M.I.) at least 2 weeks later to evaluate
intraobserver variability.
Statistical Analysis
Data are given as mean±SD. EROAs and RV/beat calculated by
using the CDVC were compared with those obtained by the EFM methods by
using linear correlation; agreement between the two measurements was
tested according to the method of Bland and Altman.19 The
EROAs were also compared with peak and mean RFRs and RV/beat by using
simple linear analysis and were compared with RFs by using
exponential regression analysis. Using trend analysis
(polynomial regression on ordered categories of
hemodynamic states), we examined each response
variable individually to determine if there were linear effects of
the hemodynamic states on each of the response
variables. In addition, because multiple points were obtained from
the same sheep and the data from each sheep were assumed to be random
samples from a larger population, a separate regression model
analysis was used to estimate the averaged correlation
coefficients of the regressions of the data obtained
echocardiographically versus data obtained
hemodynamically using the electromagnetic flow probes.
These analyses were performed by using least-squares regression
as implemented in the statistical package S-PLUS for Windows (version
3.2 supplement, StatSci Division of Mathsoft, Inc). A probability value
of <.05 was considered statistically significant.
| Results |
|---|
|
|
|---|
Echocardiography and Peak EROA
Color Doppler vena contracta images were successfully
recorded for all 29 hemodynamic states even though,
as is common clinically, all regurgitant jets were eccentric, directed
toward either the intraventricular septum (n=18) or
the anterior mitral leaflet (n=11).
The electromagnetic flow probederived EROAs varied from 0.05 to 0.46
(0.26±0.46) cm2. Good correlations were found between the
EROAs estimated from the CDVC areas and those derived by using the EFM
method (r=.91, P=.0001, SEE=0.07 cm2;
Fig 2A
). Agreement analysis using Bland and
Altman's method19 showed a tendency for slight
overestimation (Fig 2B
). A weaker but significant relationship still
existed (r=.62, P<.001) after averaging of the
separate regression analyses was used to eliminate the effects
caused by using multiple points from the same animal.
|
Relationship of the EROA to AR Severity
The results of linear regression analysis between EROAs
determined from the CDVC, peak and mean RFRs, RSVs, and the results of
the exponential correlation with EFM measurements are listed in the
Table
. The EROAs calculated by using the CDVC correlated
quite strongly with volumetric measures of the severity of AR, with the
best linear correlations derived from the relationship between the
calculated color Doppler EROAs and the peak and mean RFRs
(r=.93, .92, respectively). Weaker but significant
relationships still existed (peak RFRs: r=.78,
P=.001; mean RFRs: r=.81, P=.001;
RSVs: r=.76, P=.001) after averaging of the
separate regression analyses was used to eliminate the effects
caused by using multiple points from the same animal.
|
Regurgitant Volume
Regression analysis between the RV/beat determined by the
EFMs and the color Doppler estimates obtained by using vena
contracta measurements and CW Doppler diastolic VTIs
also demonstrated a close correlation (r=.95,
P=.0001, SEE=5.0 mL/beat; Fig 3A
). Good
correlations between peak and mean RFRs determined by
echocardiography and EFM data were also
demonstrated (peak RFR: r=.95, P=.0001, SEE=1.3
L/min; mean RFR: r=.96, P=.0001, SEE=0.47 L/min;
Figs 4A
and 5A
). Weaker but significant
relationships (RV/beat: r=.87, P=.01; peak RFR:
r=.85, P=.001; mean RFR: r=.72,
P=.001) still existed after averaging of the separate
regression analyses was used to eliminate the effects caused by
using multiple points from the same animal. The color Doppler
echocardiography estimations for all these flow
parameters also showed a tendency for overestimation of the
corresponding EFM reference results on the Bland and Altman agreement
analysis19 (Figs 3B
, 4B
, and 5B
).
|
|
|
Interobserver and Intraobserver Variability
Excellent correlations were found between the two observers for
determining the EROAs (r=.94, P=.0001, SEE=0.05
cm2) and the RV/beat (r=.95, P=.0001,
SEE=3.8 mL/beat). Interobserver absolute differences were 0.02±0.06
cm2 for EROAs and 1.2±4.1 mL/beat for RV/beat as assessed
by the CDVC method. Intraobserver variability was 0.02±0.02
cm2 for EROA determinations.
| Discussion |
|---|
|
|
|---|
Advantages of This Study Compared With Previous Studies
A variety of color Doppler flow-mapping methods have been
proposed for noninvasively estimating the severity of
AR.4 5 6 7 8 Perry et al6 studied AR jet width as a
percentage of left ventricular outflow tract diameter
compared with angiographic grading in 29 patients. They found this
measurement to be a better predictor of severity than jet area or
length. However, they did not quantify actual regurgitant flow volumes.
Holm et al17 have suggested that RSV calculated by using
the width of the flow stream combined with the CW Doppler
diastolic VTI for AR correlates favorably with angiographic
methods for grading AR. Grading AR by invasive supravalvular
aortography has been the conventional and most widely accepted standard
for evaluating the severity of AR and was used as a reference standard
in these referenced clinical studies.6 17 Angiography,
however, may be influenced by many variables, such as
arrhythmia, catheter position, amount of dye injected, x-ray
technique, and the size of the left ventricle and its systolic
and diastolic functions, including left
ventricular diastolic
compliance.20 21 22 23 Quantitative left ventricular
cineangiographic measurements in stroke volume combined with Fick or
thermodilution methods for determining forward cardiac output have also
been used for quantifying aortic regurgitant volume. These methods are
limited by the problems of deriving left ventricular volume
measurements from planar angiograms and by the inherent variability of
the thermodilution output determinations.24 25 26
Compared with previous clinical and experimental studies on the proximal jet region in AR, our study offers several advantages. We compared the color Doppler data directly with the EFM measurements of instantaneous actual RFRs. As a reference standard for the RFR, this method is more accurate than other methods.20 23 27 28 29 We used digital images of color two-dimensional Doppler for evaluating all measurements of CDVC diameter, whereas previously performed jet planimetry studies have analyzed analog video-taped images.
EROA Calculated by Using the CDVC As an Index of AR
Severity
The clinical hemodynamic assessment of
valvular regurgitation has been largely limited
to semiquantitative grading of invasive or noninvasive
parameters that have shown correlation with the regurgitant
volume.4 6 17 30 31 32 33 The assessment of EROA has been shown
to be useful for the evaluation of AR severity, since changes in EROAs
may reflect true changes in regurgitant volumes, especially for the
mitral valve.9 10 11 12 13 14 34 Determining the EROA also has
theoretical advantages. In vitro and in vivo studies indicate that the
severity of regurgitation varies with
hemodynamic status but that the EROA is not affected by
heart rate35 or driving pressure.11 34 36 Our
study shows that there are significant correlations between the EROAs
calculated by using the CDVC and hemodynamic
assessments of the severity of AR (peak and mean RFRs, RF, and RV/beat;
see Table
). Previously, the EROA has been calculated by using the
continuity equation and quantitative Doppler methods for
AR9 12 as well as by the proximal flowconvergence method
for mitral and tricuspid
regurgitation.11 13 14 27 Although
quantitative Doppler and flow-convergence methods are capable of
quantifying regurgitant flow volume, these techniques, because of their
geometric assumptions and the need for a priori knowledge of a suitable
aliasing velocity range, may be problematic to apply in
clinical settings.11 12 13 14 37 Some of the animals in the
present study had other steady-state data points that were included
in another study in which we evaluated the severity of AR by using
flow-convergence axial centerline velocity/distance
profiles.38 However, the centerline technique required
complicated computer analysis algorithms that may not be
practical for use in clinical settings. In contrast, the CDVC method
that we propose is simple and less technically demanding than these
other methods and may be more independent of loading
conditions.32 The centerline, the flow-convergence
geometric, and vena contracta methods provide alternative and/or
complementary information regarding the severity of AR.
Limitations
We used epicardial echocardiography to select
the best position for the transducer in this animal study to obtain
good alignment for color and CW Doppler interrogation of the vena
contracta and aortic regurgitant jet. In clinical situations, lower
frequencies for imaging and Doppler, such as 2.5 MHz, might be
necessary if apical views were used to image the vena contracta at
depths of
10 cm. Under such conditions, precise imaging of the vena
contracta may be difficult. However, we have found clinically that
satisfactory imaging of the vena contracta and flow convergence of AR
in most patients can be obtained by using a right parasternal view at
higher frequencies and at shorter distances. Defocusing of the
ultrasound beam or reverberation induced by the chest wall and poor
lateral resolution might lead to artifactual widening of the flow
signal and, thus, yield erroneous overestimation of the EROAs and
regurgitant volumes.36 To mitigate these problems, we used
near-field imaging (5.5±0.85 cm) with high-frequency, dynamically
focused annular array transducer imaging that focused all places
symmetrically. In clinical settings, precise alignment and optimal
imaging may not always be possible, especially when using conventional
transthoracic echo windows.39 However,
high-frequency imaging and/or atypical windows such as the
transthoracic high right parasternal view imaging aortic
regurgitant flows in the near field going away from the transducer or
transesophageal subgastric views obtained with
high-frequency devices should be capable of producing satisfactory
images in many patients. Planimetry of the cross-sectional area of the
CDVC in a short-axis view, if identified precisely and without
distortion as the smallest laminar flow area just distal to the valve
orifice, would provide direct determination of the vena contracta area
and account for situations in which the orifice is not axis
symmetric.
We studied a relatively small number of the animals and only one type of AR, which had been caused by incised, retracted aortic valve leaflets; however, this etiology of AR is clinically relevant, since one frequently encounters patients with similar anatomic pathology of the aortic valve. Additionally, regurgitant jets were directed from the central leaflet coaptation toward the anterior mitral leaflet or the interventricular septum, similar to findings in many patients. Thus, the CDVC method does appear to be applicable even for this geometrically complex regurgitant orifice. Additionally, upon examination all the orifices in our study were nearly symmetric. Asymmetric orifices, such as the slit-shaped ones that sometimes occur with bicuspid aortic valves, may require biplanar measurements to obtain the appropriate vena contracta dimensions.36 40 These potential problems can be obviated by three-dimensional reconstruction of the CDVC.41 In in vitro studies, we41 ) and others42 have demonstrated that three-dimensional reconstruction of vena contracta flow images for differently shaped orifices is possible and helpful. This technique would make vena contracta area determinations more objective and less susceptible to problems due to assumptions about geometry.
Conclusions
Our study using quantified aortic RFRs indicates that the CDVC of
AR combined with CW Doppler-determined jet velocities can be used
to estimate the EROA and to quantify the regurgitant flow volume and
flow rate. We suggest that studies of the clinical efficacy of this
method be undertaken.
| Selected Abbreviations and Acronyms |
|---|
|
| Acknowledgments |
|---|
Received April 9, 1996; revision received December 11, 1996; accepted December 16, 1996.
| References |
|---|
|
|
|---|
This article has been cited by other articles:
![]() |
K Debl, B Djavidani, S Buchner, N Heinicke, S Fredersdorf, J Haimerl, F Poschenrieder, S Feuerbach, G A J Riegger, and A Luchner Assessment of the anatomic regurgitant orifice in aortic regurgitation: a clinical magnetic resonance imaging study Heart, March 1, 2008; 94(3): e8 - e8. [Abstract] [Full Text] [PDF] |
||||
![]() |
R. Bekeredjian and P. A. Grayburn Valvular Heart Disease: Aortic Regurgitation Circulation, July 5, 2005; 112(1): 125 - 134. [Abstract] [Full Text] [PDF] |
||||
![]() |
Y. Mori, R. A. Rusk, M. Jones, X.-N. Li, T. Irvine, A. D. Zetts, and D. J. Sahn A new dynamic three-dimensional digital color doppler method for quantification of pulmonary regurgitation: validation study in an animal model J. Am. Coll. Cardiol., September 18, 2002; 40(6): 1179 - 1185. [Abstract] [Full Text] [PDF] |
||||
![]() |
D. L. Willett, S. A. Hall, M. E. Jessen, M. A. Wait, and P. A. Grayburn Assessment of aortic regurgitation by transesophageal color Doppler imaging of the vena contracta: validation against an intraoperative aortic flow probe J. Am. Coll. Cardiol., April 1, 2001; 37(5): 1450 - 1455. [Abstract] [Full Text] [PDF] |
||||
![]() |
M. Ishii, K. Hashino, G. Eto, T. Tsutsumi, W. Himeno, Y. Sugahara, H. Muta, J. Furui, T. Akagi, Y. Ito, et al. Quantitative Assessment of Severity of Ventricular Septal Defect by Three-Dimensional Reconstruction of Color Doppler-Imaged Vena Contracta and Flow Convergence Region Circulation, February 6, 2001; 103(5): 664 - 669. [Abstract] [Full Text] [PDF] |
||||
![]() |
T Irvine, X N Li, R Rusk, D Lennon, D J Sahn, and A Kenny Three dimensional colour Doppler echocardiography for the characterisation and quantification of cardiac flow events Heart, November 1, 2000; 84(90002): 2i - 6. [Full Text] |
||||
![]() |
C. M. Tribouilloy, M. Enriquez-Sarano, K. R. Bailey, A. J. Tajik, and J. B. Seward Quantification of tricuspid regurgitation by measuring the width of the vena contracta with Doppler color flow imaging: a clinical study J. Am. Coll. Cardiol., August 1, 2000; 36(2): 472 - 478. [Abstract] [Full Text] [PDF] |
||||
![]() |
C. M. Tribouilloy, M. Enriquez-Sarano, K. R. Bailey, J. B. Seward, and A. J. Tajik Assessment of Severity of Aortic Regurgitation Using the Width of the Vena Contracta : A Clinical Color Doppler Imaging Study Circulation, August 1, 2000; 102(5): 558 - 564. [Abstract] [Full Text] [PDF] |
||||
![]() |
R. Shandas, J. Kwon, and L. Valdes-Cruz A Method for Determining the Reference Effective Flow Areas for Mechanical Heart Valve Prostheses : In Vitro Validation Studies Circulation, April 25, 2000; 101(16): 1953 - 1959. [Abstract] [Full Text] [PDF] |
||||
![]() |
Y. Mori, T. Shiota, M. Jones, S. Wanitkun, T. Irvine, X. Li, A. Delabays, N. G. Pandian, and D. J. Sahn Three-Dimensional Reconstruction of the Color Doppler–Imaged Vena Contracta for Quantifying Aortic Regurgitation : Studies in a Chronic Animal Model Circulation, March 30, 1999; 99(12): 1611 - 1617. [Abstract] [Full Text] [PDF] |
||||
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
Circulation Home | Subscriptions | Archives | Feedback | Authors | Help | AHA Journals Home | Search Copyright © 1997 American Heart Association, Inc. All rights reserved. Unauthorized use prohibited. |