|
| HOME | SUBSCRIPTIONS | CURRENT ISSUE | PAST ISSUES | CARDIOSOURCE | SEARCH | HELP | FEEDBACK |
|
|
|
|||||||||
|
|||||||||||
|
J Am Coll Cardiol, 2004; 43:1654-1662, doi:10.1016/j.jacc.2003.09.066 © 2004 by the American College of Cardiology Foundation |
* Department of Cardiology, Madrid, Spain
Department of Cardiovascular Surgery, Madrid, Spain
|| Unit of Experimental Medicine and Surgery, Hospital General Universitario Gregorio Marañón, Madrid, Spain
Department of Mathematical Physics and Fluids, Facultad de Ciencias, Universidad Nacional de Educación a Distancia, Madrid, Spain
Department of Signal Theory and Communications, Universidad Carlos III de Madrid, Spain
Manuscript received May 20, 2003; revised manuscript received August 14, 2003, accepted September 29, 2003.
* Reprint requests and correspondence: Dr. Javier Bermejo, Laboratory of Echocardiography, Department of Cardiology, Hospital General Universitario Gregorio Marañón, Dr. Esquerdo 46, 28007 Madrid, Spain.
javbermejo{at}jet.es
 |
Abstract |
|---|
 Top Abstract MethodsResultsDiscussionReferences |
|---|
BACKGROUND: In the absence of obstruction, ejection IVPGs cannot be estimated by Doppler using the simplified Bernoulli equation.
METHODS: High-fidelity micromanometers were placed in the left ventricle of eight minipigs, and synchronic Doppler images and pressure signals were obtained during different hemodynamic conditions. Twenty healthy volunteers and 20 dilated cardiomyopathy patients were studied at baseline and during esmolol, dobutamine, and atropine infusion (only dobutamine in patients).
RESULTS: Excellent agreement was observed between micromanometer and Doppler methods for measuring instantaneous pressure differences among the apex, the mid-cavity, and the outflow tract (Rintraclass = 0.98, 0.81, 0.76, and 0.98 for the peak, time-to-peak, peak reverse, and time-to-peak reverse values, respectively; n = 810 beats). Error of the noninvasive method was –0.05 ± 0.25 mm Hg for the peak pressure difference. Parametrical images demonstrated that IVPGs originate mainly in the mid-ventricle and then propagate to the outflow tract. Both the magnitude and the temporal course of IVPGs were different among volunteers and patients. Inotropic interventions induced significant changes in the apex-outflow tract pressure differences in both populations, whereas atropine had no effect on IVPGs.
CONCLUSIONS: For the first time, ejection IVPGs can be accurately visualized and measured by Doppler-echocardiography. Important aspects of the dynamic interaction among myocardial performance, load mechanics, and ejection dynamics can be assessed in the clinical setting using this method.
| ||||||||||||||||
Ps) in the range of 5 to 10 mm Hg have been measured between the LV apex and the left ventricular outflow tract (LVOT) using high-fidelity micromanometers (1,4). Ejection IVPGs arise from the unsteady nature of ejective flow, from chamber geometry, and from the sequence of regional contraction (3). Therefore, ejection Ps are known to be related to the inotropic state and to LV size (2). Analysis of IVPGs is of particular interest in the clinical setting. Invasive studies have demonstrated that ejection IVPGs are responsible for a significant component of systolic afterload (3). Also, IVPGs are known to be modified by myocardial diseases (5), and the analysis of ejection IVPGs has proved to be useful in characterizing ventricular performance (3,5). However, the need for sophisticated catheterization procedures has limited further research into this field.
The simplified Bernoulli equation is an accurate method to calculate intracardiac Ps from Doppler-derived flow-velocity measurements, but is unsuitable in the absence of a restrictive orifice (6). Ejection IVPGs are mainly inertial in origin (3), and they cannot be estimated using a formula that neglects local acceleration. Recently, we and other investigators (7–10) have demonstrated that the Euler equation can be solved from color M-mode Doppler velocity data. The present study was designed with a twofold purpose: 1) in vivo validating, against high-fidelity pressure catheters, of a new noninvasive method to measure ejection IVPGs, and 2) testing the clinical applicability of the method to evaluate the effects of drug interventions on IVPGs in healthy volunteers and in patients with dilated cardiomyopathy (DCM).
|
Methods |
|---|
|
TopAbstractMethodsResultsDiscussionReferences |
|---|
The heart was exposed by median sternotomy without opening the pericardium and cradled to obtain a suitable ultrasonic window. For load manipulation, a snare was placed around the inferior vena cava, and a 20-mm valvuloplasty balloon was advanced to the infrarenal aorta. A pacing electrode was sutured to the right atrial appendage, and a needle micromanometer (Millar Instruments, Houston, Texas) was inserted epicardially inside the apex. A dual-micromanometer catheter (3-cm sensor spacing, Millar Instruments) was inserted retrogradely into the LV from the ascending aorta, placing the proximal sensor at the LVOT and the distal sensor in the mid-cavity. Micromanometers were prepared, balanced, and calibrated according to the manufacturer's specifications. To avoid drift and eliminate hydrostatic Ps (3), micromanometers were also balanced in vivo during a long diastasis following a premature beat. To ensure temporal matching of Doppler-echocardiography and pressure data, a synchronicity signal was simultaneously connected to the ultrasound scanner and the signal acquisition system. Pressure, synchronicity, and electrocardiographic (ECG) signals were digitized at 500 Hz on a Pentium III computer using custom-built amplifiers, a 16-channel A-D converter board (Keithley Instruments, Cleveland, Ohio), and virtual instrumentation software (Testpoint, Capital Equipment, Billerica, Massachusetts).
Each animal was studied at baseline and at four hemodynamic states: 1) esmolol (bolus 500 µg/kg, followed by 50 to 200 µg/kg/min in steps of 50 µg/kg/min every 10 min); 2) dobutamine (up to 10 µg/kg/min); and atrial pacing at: 3) 120 beats/min, and 4) 150 beats/min. In each state, data sets were acquired once hemodynamics were stable at baseline, during cava, and during aortic occlusion. Three consecutive beats were recorded during end-expiration apnea from each state-load condition.
Human population and clinical protocol. Twenty healthy volunteers and 20 DCM patients were studied. Volunteers were not on any medication, and cardiovascular disease was ruled out by history, clinical examination, ECG, and Doppler-echocardiography. Inclusion criteria for DCM patients were: 1) ejection fraction <30%, 2) LV volumes >2 SD normal values for age and gender, and 3) sinus rhythm. Patients with a functional class IV or prior history of sustained ventricular tachycardia were not included. In volunteers, esmolol and dobutamine (same doses as in the animal protocol) were infused, followed by atropine (0.3-mg boluses) titrated to match dobutamine heart rate. Recordings were obtained during a steady hemodynamic state, and each drug was followed by a 30-min washout period. In DCM patients, only dobutamine was infused. The study protocol was approved by the institutional ethical committee, and written informed consent was obtained from all subjects.
Echocardiographic image acquisition and analysis. A 2.0 to 3.5-MHz transducer was used on a Sequoia C-256 system (Siemens AG, Munich, Germany). Images were obtained from a posteriorly modified five-chamber apical view, using an epicardial (animals) and a transthoracic (humans) approach. Special care was taken to align the M-mode cursor with the ejective flow, visualized by two-dimensional color Doppler. Horizontal sweep was set to 150 or 200 mm/s. End-diastolic and end-systolic dimensions, shortening fraction, and LV stroke volume were calculated in the clinical study from standard M-mode, two-dimensional, and Doppler spectrograms as recommended (11).
The general method used for digital image processing of color M-mode Doppler echocardiograms has been described elsewhere (10). Briefly, if the M-mode cursor closely approximates a flow streamline, the spatio-temporal velocity distribution of a discrete blood sample is provided by the value of its corresponding pixel color: v(s,t), where v represents velocity, s represents the linear dimension of the streamline, and t is time. Thus, the color M-mode Doppler recording provides the data necessary to solve Euler's momentum equation:
 |
represents blood density. The first term in the right side of the equation designates inertial (local) acceleration, whereas the second term accounts for convective (spatial) acceleration.
Calibration and Doppler scale limits were automatically read from the image. The onset of the QRS was manually identified and defined as t = 0. Velocity was decoded, dealiased, and fitted to a tensor-product smoothing spline (smoothing parameter p = 0.004) (10). Next, partial derivatives of the velocity data were calculated as the analytical first derivative of each polynomial piece. Ejection IVPGs were displayed as color overlays representing the P between one pixel and another one located 1 cm closer to the ultrasound transducer in mm Hg/cm (Fig. 1). The locations of the LV apex and the LVOT were visually traced in the greyscale image layer, attempting to match instantaneous micromanometer positions (Fig. 1). The distance between the mid-LV and LVOT micromanometers was measured from two-dimensional echocardiographic sequences obtained at the beginning of each acquisition (slightly less than nominal separation due to catheter bending, approximately 2.6 to 2.8 cm apart). Discrete Ps were calculated by spatial integration of IVPG maps between each pair of stations, providing three validation P curves from each IVPG image. In the clinical study, the apical–LVOT P was calculated.
|
P were 0.1 ± 1.2 (Rintraclass = 0.93) and 0.2 ± 1.0 mm Hg (Rintraclass = 0.95), respectively, where Rintraclass represents the intraclass correlation coefficient.
Pressure difference curve analysis.
The P curves between apex–LVOT, apex–mid-cavity, and mid-cavity–LVOT were calculated offline by subtraction of the respective simultaneous pressure tracings. To ensure the analysis of the same beat, the temporal offset between Doppler-echocardiography and invasive data sets was obtained by cross-correlation of their synchronicity signals. The onset and the end of the ejection period were identified from the raw color-Doppler recordings as the beginning and the end of flow across the aortic valve. The following P parameters were blindly obtained for each technique, constrained to the ejection period: peak and time-to-peak, minimal and time-to-minimal, time to P inversion, and time to second inversion (slightly after end ejection). In the clinical study, both the value and the time of peak IVPG were also measured from the parametrical map. Temporal parameters were normalized to ejection duration in the clinical study.
Statistical analysis.
The validation analysis was performed on a single-beat basis, whereas results of three consecutive cycles were averaged in the clinical study. Quantitative variables are expressed as the mean ± SD. The agreement between techniques was assessed calculating the mean ± SD of error, Bland-Altman charts, and the intraclass correlation coefficient (Rintraclass). In the clinical study, differences among phases were assessed by analysis of variance accounting for repeated-measures, followed by Dunnet's contrasts. Unpaired t tests were used for comparisons between healthy volunteers and patients. Variables related to peak P were assessed calculating Pearson's correlation coefficients, followed by multivariate linear regression analysis with variable selection based on Alkaike's AIC criterion. A p value <0.05 was considered to be significant.
|
Results |
|---|
|
TopAbstractMethodsResultsDiscussionReferences |
|---|
Ps were 0.7 ± 0.2, 2.5 ± 1.5, and 2.9 ± 1.7 mm Hg, for the apex–mid-LV, mid-LV–LVOT, and apex–LVOT locations, respectively. The P curves obtained by Doppler-echocardiography closely matched invasive signals, and a low-pass filter effect was observed in noninvasive tracings (Fig. 1). The accuracy of Doppler-echocardiography P curve parameters was excellent (Table 1, Fig. 2). Beat-to-beat variability (animal data set, baseline states of all phases, n = 66, three consecutive beats) for peak ejection P values was 0.6 ± 0.5 and 0.5 ± 0.5 mm Hg, for Doppler-echocardiography and micromanometer methods, respectively. The phase lag between methods was 7 ± 25 ms. Ejection Ps adopted a characteristic configuration (Figs. 1 and 3 ) showing an early peak, a slow decrease, a mid-late ejection P reversal, and, finally, a late reverse peak. Ejection IVPGs were caused mainly by inertial forces, whereas convective P curves did not resemble micromanometer tracings (Fig. 3).
|
|
|
P, and lower peak-reverse P values, mainly due to lower convective forces. The time to peak P and the time to P inversion occurred earlier during ejection in DCM patients than in normal subjects, whereas the time to second inversion occurred later. Dobutamine significantly increased all absolute values of IVPG and P. Conversely, peak positive values were decreased by esmolol in normal subjects. Atropine did not modify either IVPG or P parameters (Table 2). Peak Ps correlated with heart rate, systolic blood pressure, and, most significantly, with stroke volume (Table 3).
|
|
|
|
|
Discussion |
|---|
|
TopAbstractMethodsResultsDiscussionReferences |
|---|
Pressure gradients assessed by color M-mode Doppler.
The possibility of calculating intracardiac Ps from color M-mode Doppler images was first demonstrated by Greenberg et al. (7). By solving the unsteady Bernoulli equation, these investigators accurately measured diastolic Ps across a normal mitral valve in an open-chest animal model (7), and in patients undergoing coronary artery bypass grafting (9). Later, the same group validated a method for estimating diastolic intraventricular Ps, based on the solution of the Euler equation (8). Our method, proposed almost simultaneously (10), introduces some potential improvements. First, velocity derivatives are solved analytically, increasing the robustness of mathematical calculations. Second, the effects of cardiac translation can be partially corrected by manually locating the measuring stations throughout the entire sequence. And third, not only P curves, but also parametric IVPGs images are provided, showing the spatio-temporal distribution of intracardiac pressure forces. We believe this type of representation provides useful information. A previous study (2) and ours demonstrate that about 25% to 30% of ejection IVPGs hold between the apex and the mid-cavity stations. These apical ejection IVPGs have never been characterized in humans, and can now be visualized and measured with reasonable accuracy (Table 1, Figs. 4 and 5).
A number of implicit assumptions are present in Euler's one-dimensional simplification of ejection flow-dynamics: 1) viscous frictional effects are neglible in the ejection field and confined to very thin boundary layers; 2) the main core of LV outflow is irrotational; and 3) there is an absence of vorticity in the LV ejection flow. These mechanical aspects of ventricular ejection, to date incompletely addressed in vivo (12), are supported by our validation results.
Methods based on solving fluid-dynamic differential equations, such as the one we propose, are accurate to calculate regional differences of pressure within and between the cardiac chambers. Absolute local pressure curves can only be inferred if instantaneous pressure at a given position is either known or assumed. Furthermore, to obtain absolute pressure from flow velocity accurately, not only local but also global inertial forces (those required to accelerate the total blood within the ventricle) would also have to be accounted for (5,13).
Ejection IVPGs in health and disease.
Laszt and Müller described the existence of ejection intraventricular regional Ps more than 50 years ago (14). However, ejection Ps could not be well characterized until the introduction of high-fidelity solid-state pressure catheters (15). With this tool, ejection intraventricular Ps were described in dogs, and the relationship with myocardial contractility was suggested (2). Also, intraventricular ejection Ps were studied invasively in a small number of normal subjects (1). Remarkably, both the magnitude and the shape of P curves obtained in our study are identical to the micromanometer tracings provided by these studies (1,3). Although intraventricular Ps have never been analyzed in humans undergoing pharmacological interventions, our findings agree with previous invasive animal studies: peak ejection Ps are increased by beta-agonists (2,16), and reduced by beta-blocking agents (2,16).
Invasive assessment of intraventricular Ps has helped clarify the hemodynamics of obstructive and nonobstructive forms of hypertrophic cardiomyopathy (4). In aortic valve stenosis, ejection intraventricular Ps are responsible for about half the total transvalvular P, and they are a major determinant of cardiac output (17,18). However, to our knowledge, ejection intraventricular Ps have never been studied invasively in patients with DCM. Isaaz and Pasipoularides (5) used a noninvasive method to derive ejection Ps from manually digitized pulsed-wave Doppler spectrograms in patients with DCM. Their study also demonstrated that, although theoretically greater disproportion between the mid-LV and the LVOT should increase convective IVPGs in dilated ventricles (3), reduced ejection velocities due to systolic dysfunction cause convective forces to be lower than in normal hearts (5). For this reason, the time to P reversal should take place earlier in DCM than in normal subjects. Our results in the DCM population are fully concordant with these observations (Table 2).
Clinical applications. The estimation of systolic IVPGs has important implications for the assessment of ventricular load. Our study validates a new noninvasive method to quantify intrinsic afterload (19) by Doppler-echocardiography. The results of our and previous studies (2,5) indicate that the intrinsic load adds about 10% to the extrinsic load in normal subjects, a value that is close to the oscillatory component of aortic input impedance. Interestingly, the oscillatory component of longitudinal intraventricular systolic impedance can also be assessed using our method, and preliminary results suggest it might not be negligible (20).
The use of the temporal derivative of LV pressure (dP/dt) as an index of myocardial contractility is based on the diagnostic value of initial impulsive forces (21). A computer fluid-dynamic simulation has suggested a theoretical advantage of peak intraventricular P and flow acceleration over peak dP/dt (22). Flow acceleration has recently been shown to provide a robust estimator of peak systolic elastance (23). Similarly, the analysis of ejection IVPGs may be helpful to assess contractility (24).
Reversed P during late ejection is the consequence of the kinetic energy of blood due to the momentum imparted to it earlier in systole (21,25). As shown in our study (Figs. 1, 3, and 4), negative ejection gradients hold up to the isovolumic relaxation period. During isovolumic relaxation a negative (suction) P has been described between the LV apex and the LVOT (26). Therefore, the temporal extent of diastolic-reversed IVPGs could be useful to determine the rate of myocardial relaxation (24).
Study limitations. Ejection velocities measured using Doppler represent the combination of intraventricular acceleration within the heart, plus the effect of ventricular translation within the chest. Because these two components are of opposite direction, Doppler-derived velocities are lower than net intraventricular velocities (27). Although this would lead to underestimation of inertial acceleration, it has been recently demonstrated that longitudinal displacement of the center of mass of the LV is in the range of ±5 cm/s (28). Therefore, we believe this effect should not lead to significant inaccuracy either in open or in closed-chest setups.
A second potential limitation is that the validation study was not performed in humans, owing to the need of multisite micromanometer instrumentation. Because geometrical properties of the pig's LV closely resemble those of humans, we do not anticipate relevant species-related differences. A third limitation is related to LV outflow profile. In this study, outflow velocities were obtained across the center of the LVOT. However, it is known that outflow velocity is slightly skewed (29). According to Euler's equation, regional variations in outflow profile should modify flow velocities and IVPGs in parallel. For this reason, special care was taken in the study to match micromanometer and echo-Doppler measuring stations. Nevertheless, the skewed outflow profile must be recognized as a source of limited consistency and reproducibility of all methods that measure ejection IVPGs. Finally, healthy volunteers and DCM patient groups were not matched for age; the comparison of hemodynamic findings between both groups should therefore be interpreted with caution.
Conclusions. This study validates, for the first time, a fully noninvasive method to assess the dynamic interaction among myocardial performance, load mechanics, and intraventricular fluid dynamics in the clinical setting. Although the method is based on offline processing of digital color M-mode Doppler images, it could be incorporated into the analysis software of future ultrasound scanners.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
|
References |
|---|
|
TopAbstractMethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  C. Cortina, J. Bermejo, R. Yotti, M. M. Desco, D. Rodriguez-Perez, J. C. Antoranz, J. L. Rojo-Alvarez, D. Garcia, M. A. Garcia-Fernandez, and F. Fernandez-Aviles Noninvasive Assessment of the Right Ventricular Filling Pressure Gradient Circulation, August 28, 2007; 116(9): 1015 - 1023. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. Yotti, J. Bermejo, J. C. Antoranz, M. M. Desco, C. Cortina, J. L. Rojo-Alvarez, C. Allue, L. Martin, M. Moreno, J. A. Serrano, et al. A Noninvasive Method for Assessing Impaired Diastolic Suction in Patients With Dilated Cardiomyopathy Circulation, November 8, 2005; 112(19): 2921 - 2929. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. D. Thomas and Z. B. Popovic Intraventricular Pressure Differences: A New Window Into Cardiac Function Circulation, September 20, 2005; 112(12): 1684 - 1686. [Full Text] [PDF]
|
|
|
|
|
|
R. Yotti, J. Bermejo, M. M. Desco, J. C. Antoranz, J. L. Rojo-Alvarez, C. Cortina, C. Allue, H. Rodriguez-Abella, M. Moreno, and M. A. Garcia-Fernandez Doppler-Derived Ejection Intraventricular Pressure Gradients Provide a Reliable Assessment of Left Ventricular Systolic Chamber Function Circulation, September 20, 2005; 112(12): 1771 - 1779. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. E. Weyman The year in echocardiography J. Am. Coll. Cardiol., February 1, 2005; 45(3): 448 - 455. [Full Text] [PDF]
|
|
|
|
|
|
J. Bermejo, J. L. Rojo-Alvarez, J. C. Antoranz, M. Abel, I. G. Burwash, R. Yotti, M. Moreno, M. A. Garcia-Fernandez, K. G. Lehmann, and C. M. Otto Estimation of the End of Ejection in Aortic Stenosis: An Unreported Source of Error in the Invasive Assessment of Severity Circulation, August 31, 2004; 110(9): 1114 - 1120. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | SUBSCRIPTIONS | CURRENT ISSUE | PAST ISSUES | CARDIOSOURCE | SEARCH | HELP | FEEDBACK |
p/
