| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Technical Developments |
1 From the Department of Cardiovascular Radiology, Hôpital Broussais, 96 rue Didot, 75014 Paris, France (E.M., O.J., J.C.G.); Institut National de la Santé et de la Recherche Médicale (INSERM) U494, Centre Hospitalier Universitaire (CHU) Pitié-Salpêtrière, Paris, France (E.M., O.J.); Unité de Recherche Associée (URA), Centre National de la Recherche Scientifique (CNRS) 2212, Centre Iuter Etablissement de Résonance Magnétique (CIERM), Hôpital de Bicêtre, Le Kremlin Bicêtre, France (J.P.T., J.B.); and the Department of Respiratory Disease, Hôpital Antoine Béclère, Clamart, France (G.S.). From the 1998 RSNA scientific assembly. Received May 29, 1998; revision requested July 16; final revision received November 20; accepted February 22, 1999. Address reprint requests to E.M. (e-mail: mousseaux@hbroussais.fr).
 |
Abstract |
|---|
 TOP Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Index terms: Hypertension, pulmonary, 564.78 • Magnetic resonance (MR), phase imaging, 564.12144 • Pulmonary arteries, abnormalities, 564.78 • Pulmonary arteries, flow dynamics, 564.78
|
Introduction |
|---|
|
TOPAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
It is possible to directly and accurately measure stroke volume of the right and left ventricles and the corresponding cardiac output with velocity-encoded magnetic resonance (MR) imaging in the main pulmonary artery or in the ascending aorta (3–9). At velocity-encoded MR imaging in patients with pulmonary hypertension, pulmonary arterial flow measurements have indicated very inhomogeneous velocity profiles, large volumes of retrograde flow, and low distensibility of the main pulmonary artery (5,7,10). Results of velocity-profile analyses within the pulmonary artery and throughout the cardiac cycle at velocity-encoded MR imaging also suggest that quantification of selected parameters of blood flow may be used as a noninvasive method for estimating PVR.
The natural logarithm of the percentage of pulmonary retrograde flow obtained at velocity-encoded MR imaging was found to be proportional to PVR (5). As previously found in Doppler studies (11,12), the acceleration time, defined in velocity-encoded MR imaging as the time from the onset of flow to the peak velocity, was much shorter in patients with primary pulmonary hypertension than it was in healthy volunteers (7), and it was inversely correlated with mean pulmonary arterial pressure (7,13). In other Doppler studies (14,15), however, the acceleration time was found to be a poor predictor of pulmonary arterial pressure or PVR. During ejection, the right ventricle generates pulsatile pressure and flow waves that interact with and are modified by characteristics of the distensible pulmonary arterial circulation.
The present study was performed to describe and assess the accuracy of indexes of PVR measurements obtained at velocity-encoded MR imaging on the basis of estimations of the instantaneous main pulmonary arterial blood flow and wave form morphology. The accuracy of estimates of cardiac output at velocity-encoded MR imaging was also evaluated and used to assess the measurements of instantaneous pulmonary arterial blood flow. Thus, the present study was designed to determine whether velocity-encoded MR imaging within a single cross-sectional area of the pulmonary artery can provide reliable measurements of cardiac output and PVR, which are recognized as prognostic hemodynamic factors in patients with pulmonary hypertension.
Cardiac output and various indexes derived from pulmonary arterial blood flow were estimated in patients by means of velocity-encoded MR imaging, and results were compared with cardiac output and PVR values obtained by means of catheterization of the right side of the heart. The indexes derived at velocity-encoded MR imaging were also estimated within the main pulmonary artery and the ascending aorta in healthy volunteers, and those data were compared with the findings obtained in patients with and patients without pulmonary hypertension.
|
Materials and Methods |
|---|
|
TOPAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Volunteers.—Ten healthy volunteers without history of cardiovascular events (four women, six men; age range, 20–33 years; mean age, 25 years) were examined with velocity-encoded MR imaging.
Patients.—Nineteen patients (eight women, 11 men; age range, 20–65 years; mean age, 38 years) gave their informed consent for right-sided heart catheterization and MR imaging; the procedures were performed within 48 hours of each other (mean, 28 hours). No change in drug treatment and no clinically important evolution of the disease occurred in the interval between the two examinations. Indications for cardiac catheterization were the assessment of primary (n = 10) and secondary (n = 5) pulmonary hypertension and the follow-up of cardiac transplants (n = 4). The patients were assigned to one of two groups based on the PVR findings after right-sided heart catheterization. Group 1 patients (n = 12) had a PVR of more than 6 mm Hg · min · L-1; all 10 patients with primary pulmonary hypertension were in this group. Group 2 patients (n = 7) had a PVR of less than 6 mm Hg · min · L-1.
MR Flow Techniques
MR imaging was performed on a 1.5-T system (Signa; GE Medical Systems, Milwaukee, Wis). In the healthy volunteers, two series of localizer images were obtained in the axial and coronal planes. Then two images were acquired for velocity measurements. The first was in a double-oblique section perpendicular to the direction of the ascending aorta at the level of the pulmonary bifurcation, and the second was in a double-oblique section perpendicular to the main pulmonary artery, 10–15 mm above the pulmonary valves. Only one set of velocity measurements was acquired for each patient, perpendicular to the main pulmonary artery.
In the present study, the three-dimensional Fourier method for velocity-encoded MR imaging was achieved by means of four velocity-encoded gradient steps in the direction of the section-selection gradient followed by zero-filling interpolation along the 256 points of the velocity dimension around the four central planes of the acquisition (16). The parameters of the sequence included repetition time of 21 msec and echo time of 12 msec, a field of view of 32 cm, an acquisition matrix of 128 x 256 visualized in 256 x 256, and a constant 60° flip angle. The velocity-encoding gradient was adjusted so that velocities from 1.5 to -1.5 m/sec were measured without aliasing. Depending on the heart rate, 25–35 time frames were acquired with electrocardiographic gating during the first 80% of the cardiac cycle. Each velocity-encoded MR study was acquired in 4–8 minutes, according to the heart rate.
Quantitative Analysis of Flow at Velocity-encoded MR Imaging
The reconstructed velocity images were color coded. The color scale range was from 1.5 to -1.5 m/sec during systole and from 0.35 to -0.35 m/sec during diastole, to optimize the lowest velocities and provide optimal differentiation between structures on the velocity image. The observer was provided with simultaneous displays of the magnitude and corresponding velocity images for each phase of the cardiac cycle. Use of a semiautomated process helped the operator to precisely superimpose a region of interest on the vessel section at each phase of the cardiac cycle.
Stroke volume and output calculations.—In the ascending aorta and main pulmonary artery, left and right ventricular stroke volumes were estimated by means of integration of the flow velocity over the vessel section and then over the cardiac cycle. The output of both ventricles was then estimated by multiplying the corresponding stroke volume by the heart rate. The heart rate was averaged for each series over the 512 cardiac cycles of the acquisition period.
Calculation of hemodynamic indexes.—The curve of pulmonary arterial flow rate as a function of time over the cardiac cycle was used to estimate the antegrade and retrograde flow volumes and the percentage of retrograde flow. Thus, the pulmonary regurgitation fraction was defined as the ratio of the retrograde to the antegrade flow volumes. The acceleration time was defined as the time from the onset of flow to the peak velocity (Fig 1). Three other criteria were also analyzed: the maximal change in flow rate during ejection, defined as the maximal value of the ascending slope of the flow rate; the acceleration volume, estimated by integrating the flow rate of the vessel from the onset of ejection flow to the peak velocity; and the ratio of these two criteria.
|
|
|
Statistical Analysis
All velocity-encoded MR imaging measurements were performed independently by two operators (E.M., J.P.T.) and repeated by one operator (E.M.) 2 months later to assess the inter- and intraobserver agreement, respectively. Each operator estimated velocity-encoded MR imaging measurements blindly, and neither operator was aware of the right-sided heart catheterization results. The estimates of left and right ventricular stroke volume and cardiac output obtained at velocity-encoded MR imaging and thermodilution were compared by means of linear regression analysis and calculation of the standard error of the estimate (SEE). The relationship between the mean and the difference of measurements was used to complement the linear regression analyses. Paired Student t tests were also performed for such comparisons. Linear regression analysis was used to compare PVR measurements obtained at right-sided heart catheterization and the indexes of pulmonary blood flow derived at velocity-encoded MR imaging. The flow-derived indexes obtained at velocity-encoded MR imaging in the pulmonary arteries in groups 1 and 2 and healthy volunteers and within the ascending aorta in the healthy volunteers were compared by means of analysis of variance. Index values were given as the mean plus or minus one SEE. A P value less than .05 was considered significant.
|
Results |
|---|
|
TOPAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
|
|
|
|
|
|
The best hemodynamic parameter for differentiating between groups 1 and 2 was the ratio of maximal change in flow rate during ejection to acceleration volume in the pulmonary artery. Figure 3c shows that data for only one group 1 patient overlapped those for the group 2 patients with high PVR. Estimates of the ratio in the ascending aorta of healthy volunteers were also between those in the pulmonary artery for groups 1 and 2. Figure 4 illustrates the correlation between the ratios measured by the two observers. The mean difference between observers was 12 sec-2 ± 39, which was not statistically different from zero. Furthermore, no significant relationship was found between mean measurements of the two observers and the difference between these measurements. The correlation coefficient and slope of the curve for intraobserver reproducibility were 0.99 and 1.03, respectively, and the SEE was 20 sec-2.
|
|
Discussion |
|---|
|
TOPAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
6 mm Hg · min · L-1) from patients with nearly normal PVR or healthy volunteers. Few studies of measurement of cardiac output with velocity-encoded MR imaging have provided results close to those obtained with well-accepted invasive techniques such as thermodilution (6–9,17) or use of the Fick principle (8,9). Small differences between measurements of right-sided cardiac output or stroke volume with velocity-encoded MR imaging or thermodilution were found in the present study; these differences were similar to those in the previous studies. The differences may be due to the absence of simultaneous measurements, which emphasizes the spontaneous variability of cardiac output in patients (18,19) and the known limitations of the thermodilution technique for assessment of low cardiac output (19). An advantage of velocity-encoded MR imaging over thermodilution for estimation of cardiac output is that it is noninvasive and depends less on changes in stroke volume from one cardiac cycle to another (in the present study, measurements were averaged over 512 consecutive cardiac cycles).
The estimates of pulmonary flow volume and flow-time curves with velocity-encoded MR imaging gave hemodynamic parameters that are inversely or directly proportional to PVR. The decrease in acceleration time was first proposed as a criterion for pulmonary arterial hypertension or increased PVR in Doppler studies (11,20,21). The increased magnitude and earliness of the reflection wave may cause premature closure of the pulmonary valve in patients with high chronic pulmonary arterial pressure. Results at velocity-encoded MR imaging have shown a good inverse correlation (r = 0.90) between mean pulmonary arterial pressure and acceleration time (7) in patients with severe primary pulmonary hypertension and mean pulmonary arterial pressures of 40–83 mm Hg. Patients without primary pulmonary hypertension were also included in the present study, and we found a significant correlation between acceleration time and PVR (r = 0.65). The pulmonary acceleration time was also significantly lower in patients with high PVR than in those with nearly normal PVR or in healthy volunteers.
The acceleration volume was inversely correlated with PVR (r = 0.78), and the correlation with this parameter was better than that for the relation between acceleration time and PVR. Patients with high PVR were also clearly distinguished from those with low PVR on the basis of estimates of acceleration volume. This parameter emphasizes the shorter acceleration time and the lower pulmonary blood flow throughout this period in patients with pulmonary hypertension. The acceleration volume, which is influenced by the right ventricular contractile forces, may represent the initial mass of pulsatile blood encountering the elastic properties of the pulmonary arterial tree and opposing inertial forces and reflection waves. This resistance to ejected blood flow in cases of pulmonary hypertension leads to premature peak systolic blood flow and early retrograde flow, which begins just after peak blood flow and increases until end systole (5,7,10,22). These flow profile characteristics during the cardiac cycle were also found in the present study. However, unlike in previous velocity-encoded MR imaging studies, there was a poor correlation between retrograde flow and PVR.
Maximum acceleration, estimated as the first temporal derivative of Doppler velocity, has been described as a good inotropic index of the left ventricle with low load and heart-rate dependence when blood velocity is analyzed within the ascending aorta (23–27) or of the right ventricle when blood velocity is analyzed within the pulmonary artery (28). However, acceleration also depends on the distensibility of the vessel at its origin from the heart and on changes in the velocity profile during the early phase of ejection. Both the distensibility of the vessel and the velocity profile changes can be taken into account on the basis of maximal change in flow rate during ejection rather than on acceleration. Because PVR has a high dependence on volume load, the lack of correlation between the maximal change in flow rate during ejection for pulmonary flow and PVR confirms the low dependence of this parameter on volume load. Maximal change in flow rate during ejection has also been used to normalize the acceleration volume by means of an inotropic index.
In this study, the ratio of maximal change in flow rate during ejection to acceleration volume was well correlated with PVR (r = 0.89). It is thus a suitable parameter that is directly related to PVR. The ratio was much higher in group 1 than in group 2 (P < .001). Only one group 2 patient, with a PVR of 5 mm Hg · min · L-1 (patient 13 in Table 2), had a ratio of maximal change in flow rate during ejection to acceleration volume (387 sec-2) that overlapped the values for group 1. Acceleration volume and the ratio of maximal change in flow rate during ejection to acceleration volume were significantly lower in the ascending aorta flow than in the pulmonary arterial flow in healthy volunteers, even though the right and the left ventricular stroke volumes were similar, whereas the aortic maximal change in flow rate during ejection was higher. The ratios of maximal change in flow rate during ejection to acceleration volume in the ascending aorta in healthy volunteers were also close to the values for group 1 patients with high PVR.
Although systemic resistance was not assessed in our healthy volunteers younger than 30 years, it is likely to be close to the PVR of group 1 patients (13 mm Hg · min · L-1 ± 5). Thus, an increase in PVR gives rise to forces opposing the pulsatile ejected volume in the pulmonary artery that produce the acceleration time, acceleration volume, and ratio of maximal change in flow rate during ejection to acceleration volume in the ascending phase of the aortic inflow in healthy volunteers. Further experimental studies with simultaneous measurements of input impedance, PVR, and flow measurement within the pulmonary artery with velocity-encoded MR imaging will provide a better understanding of the meanings of acceleration volume and the ratio of maximal change in flow rate during ejection to acceleration volume. These hemodynamic parameters at velocity-encoded MR imaging are mainly reflections of the right ventricular afterload, however, which is accepted as a good indicator of poor outcome in patients with primary pulmonary hypertension (29).
Intra- and interobserver reproducibilities were good in the present study for measurements of cardiac output and stroke volume and index estimates derived from the flow-rate curve throughout the cardiac cycle at velocity-encoded MR imaging. This can be attributed to the semiautomatic analysis that was used. High flow signal intensities were depicted on magnitude images of great vessels obtained in patients with high velocities during systole, owing to the time-of-flight effect and very homogeneous velocity distribution. On either magnitude or velocity images, use of a thresholding method can help observers define the vessels more independently than they can by simply placing a circular region of interest as close as possible to the vascular wall. The Fourier method used for velocity-encoded MR imaging gave velocity values close to the aliasing velocity that were as accurate as those for velocities close to zero.
In conclusion, the mechanical relationship between the right ventricle and the pulmonary circulation has not been extensively studied, to our knowledge. The paucity of studies is essentially due to the need to use invasive techniques, but these cannot be justified ethically for the majority of patients. Estimates of pulmonary blood flow at velocity-encoded MR imaging should allow assessment of cardiac output and blood-flow–derived indexes of PVR in patients with right-side heart disease or primary or secondary pulmonary hypertension. This completely noninvasive approach should increase our understanding of the pathophysiologic changes affecting the pulmonary circulation of patients with various pulmonary diseases and should help in the follow-up of hemodynamic changes associated with medical therapies such as long-term vasodilator treatment in patients with primary pulmonary hypertension. Additional studies with different MR imaging systems, different velocity-encoded MR techniques, and inclusion of the appropriate spectrum of patients with pulmonary hypertension should be undertaken to validate the present results.
|
Footnotes |
|---|
Author contributions: Guarantors of integrity of entire study, J.B., J.C.G.; study concepts, E.M., O.J., G.S.; study design, E.M., J.P.T.; definition of intellectual content, E.M., J.P.T., O.J.; literature research, J.P.T., E.M.; clinical studies, G.S., E.M.; experimental studies, J.P.T., E.M.; data acquisition, J.P.T., E.M.; data analysis, E.M., J.P.T., G.S.; statistical analysis, E.M., J.P.T.; manuscript preparation, E.M., J.P.T., O.J.; manuscript editing, E.M., O.J.; manuscript review, all authors.
|
References |
|---|
|
TOPAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  G. Reiter, U. Reiter, G. Kovacs, B. Kainz, K. Schmidt, R. Maier, H. Olschewski, and R. Rienmueller Magnetic Resonance-Derived 3-Dimensional Blood Flow Patterns in the Main Pulmonary Artery as a Marker of Pulmonary Hypertension and a Measure of Elevated Mean Pulmonary Arterial Pressure Circ Cardiovasc Imaging, July 1, 2008; 1(1): 23 - 30. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. Sanz, P. Kuschnir, T. Rius, R. Salguero, R. Sulica, A. J. Einstein, S. Dellegrottaglie, V. Fuster, S. Rajagopalan, and M. Poon Pulmonary Arterial Hypertension: Noninvasive Detection with Phase-Contrast MR Imaging Radiology, April 1, 2007; 243(1): 70 - 79. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. Jardim, C. E. Rochitte, M. Humbert, G. Rubenfeld, D. Jasinowodolinski, C. R. R. Carvalho, and R. Souza Pulmonary artery distensibility in pulmonary arterial hypertension: an MRI pilot study Eur. Respir. J., March 1, 2007; 29(3): 476 - 481. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 Y. Ohno, H. Hatabu, K. Murase, T. Higashino, M. Nogami, T. Yoshikawa, and K. Sugimura Primary Pulmonary Hypertension: 3D Dynamic Perfusion MRI for Quantitative Analysis of Regional Pulmonary Perfusion Am. J. Roentgenol., January 1, 2007; 188(1): 48 - 56. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. Ley, C. Fink, M. Puderbach, J. Zaporozhan, C. Plathow, M. Eichinger, W. Hosch, K.-F. Kreitner, and H.-U. Kauczor MRI Measurement of the hemodynamics of the pulmonary and systemic arterial circulation: influence of breathing maneuvers. Am. J. Roentgenol., August 1, 2006; 187(2): 439 - 444. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K. G. Blyth, B. A. Groenning, T. N. Martin, J. E. Foster, P. B. Mark, H. J. Dargie, and A. J. Peacock Contrast enhanced-cardiovascular magnetic resonance imaging in patients with pulmonary hypertension Eur. Heart J., October 1, 2005; 26(19): 1993 - 1999. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
K.-F. J. Kreitner, S. Ley, H.-U. Kauczor, E. Mayer, T. Kramm, M. B. Pitton, F. Krummenauer, and M. Thelen Chronic Thromboembolic Pulmonary Hypertension: Pre- and Postoperative Assessment with Breath-hold MR Imaging Techniques Radiology, August 1, 2004; 232(2): 535 - 543. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. McGoon, D. Gutterman, V. Steen, R. Barst, D. C. McCrory, T. A. Fortin, and J. E. Loyd Screening, Early Detection, and Diagnosis of Pulmonary Arterial Hypertension: ACCP Evidence-Based Clinical Practice Guidelines Chest, July 1, 2004; 126(1_suppl): 14S - 34S. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 E. Laffon, C. Vallet, V. Bernard, M. Montaudon, D. Ducassou, F. Laurent, and R. Marthan A computed method for noninvasive MRI assessment of pulmonary arterial hypertension J Appl Physiol, February 1, 2004; 96(2): 463 - 468. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
N.L. Muller Computed tomography and magnetic resonance imaging: past, present and future Eur. Respir. J., February 1, 2002; 19(35_suppl): 3S - 12s. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
E. Laffon, F. Laurent, V. Bernard, L. De Boucaud, D. Ducassou, and R. Marthan Noninvasive assessment of pulmonary arterial hypertension by MR phase-mapping method J Appl Physiol, June 1, 2001; 90(6): 2197 - 2202. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. T. Marcus, A. Vonk Noordegraaf, R. J. Roeleveld, P. E. Postmus, R. M. Heethaar, A. C. Van Rossum, and A. Boonstra Impaired Left Ventricular Filling Due to Right Ventricular Pressure Overload in Primary Pulmonary Hypertension : Noninvasive Monitoring Using MRI Chest, June 1, 2001; 119(6): 1761 - 1765. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| RADIOLOGY | RADIOGRAPHICS | RSNA JOURNALS ONLINE |
) and one with high PVR (•). In (b) a patient with normal PVR and (c) a patient with high PVR, double-oblique images obtained with electrocardiography-gated velocity-encoded MR imaging perpendicular to the main pulmonary artery, 10-15 mm above the pulmonary valves, depict the corresponding pulmonary flow patterns at 133 msec (top left), 181 msec (top right), 253 msec (bottom left), and 325 msec (bottom right). The patients have the same heart rate. These images are a result of superimposition of a velocity-encoded region of interest (color coded in blue or red) in the cross-sectioned main pulmonary artery on magnitude images (color coded in black or white). Note the lower acceleration volume in c—associated with the short acceleration time, increased cross-sectional area, low peak antegrade flow (blue), and large retrograde flow (red) beginning soon after the peak—compared to that in b.
