| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Cardiac Imaging |
1 From the Department of Radiology, University of California, San Francisco, 505 Parnassus Ave, Rm L-308, San Francisco, CA 94143-0628 (G.K.L., N.W., M.S., G.P.R., M.Y., P.A.A., C.B.H.); Altos Cardiovascular, Los Altos, Calif (D.C.); and Novartis Pharmaceuticals, East Hanover, NJ (M.B.). Received January 14, 2002; revision requested March 5; final revision received August 13; accepted August 26. Supported in part by Novartis Pharmaceuticals, East Hanover, NJ. G.K.L. supported in part by a scholarship from the University Hospital Eppendorf, Hamburg, Germany. N.W. supported by a scholarship from the Max-Kade Foundation, New York, NY. Address correspondence to C.B.H. (e-mail: charles.higgins@radiology.ucsf.edu).
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODS RESULTS DISCUSSIONREFERENCES |
|---|
MATERIALS AND METHODS: MR measurements were performed in 10 consecutive patients with chronic heart failure due to coronary artery disease and in 10 volunteers. Global LV perfusion was quantified by measuring coronary sinus flow in an oblique imaging plane perpendicular to the coronary sinus with non–breath-hold VEC MR imaging. LV mass was measured by means of cine imaging that encompassed the heart. LV perfusion was calculated from coronary sinus flow and mass. CFR was measured from LV perfusion at rest and that after infusion of dipyridamole. Analysis of covariance was used to determine differences between groups. Differences within groups were analyzed by means of the Student t test for paired data. Regression analysis was used to determine correlation between CFR and LV ejection fraction.
RESULTS: At rest, LV perfusion was not significantly different in patients with chronic heart failure (0.46 mL/min/g ± 0.19) and volunteers (0.52 mL/min/g ± 0.21, P = .54). After administration of dipyridamole, LV perfusion was less than half in patients with chronic heart failure compared with that in volunteers (1.07 mL/min/g ± 0.64 vs 2.19 mL/min/g ± 0.98) (P = .03). CFR was severely reduced in patients with chronic heart failure compared with that in volunteers (2.3 ± 0.9 vs 4.2 ± 1.5, P = .01). A moderate but significant correlation was found between CFR and LV ejection fraction (r = 0.54, P = .02)
CONCLUSION: Combined cine and VEC MR imaging revealed that patients with chronic heart failure have normal LV perfusion at rest but severely depressed LV perfusion after vasodilation. Impaired CFR may contribute to progressive decline in LV function in patients with chronic heart failure.
© RSNA, 2003
Index terms: Heart, flow dynamics, 50.77, 524.12144, 5475.12144, 58.12144 • Heart, MR, 524.12144, 5475.12144, 58.12144 • Heart, ventricles, 511.12144 • Magnetic resonance (MR), perfusion study, 524.12144, 5475.12144, 58.12144
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTS DISCUSSIONREFERENCES |
|---|
Global LV perfusion can be quantified by measuring coronary sinus flow because the coronary sinus drains a large portion of the LV myocardium (7). Findings in recent studies show the potential of velocity-encoded cine (VEC) magnetic resonance (MR) imaging to quantify blood flow in small vessels (8) and to noninvasively measure coronary sinus flow (9,10). In a pulsatile flow phantom study, Arheden et al (8) determined the value of VEC MR imaging in a moving vessel that simulated cardiac motion. They found that VEC MR imaging is accurate in measuring average flow and flow profiles in vessels as small as 6 mm in diameter. Lund et al (9) validated the accuracy of this MR method to quantify coronary sinus flow by using flow probes in dogs. They found a strong correlation between VEC MR measurement of coronary sinus flow and that of the left anterior descending and circumflex coronary arteries. These data indicate that coronary sinus flow at VEC MR imaging represents an excellent surrogate measure of global LV perfusion. Schwitter et al (10) compared coronary sinus flow at VEC MR imaging with measurements of LV perfusion at positron emission tomography (PET) in humans. Good correlation and agreement were found between the two techniques, which emphasizes the value of MR imaging in the quantification of LV perfusion.
The purpose of the current study was to quantify and compare global LV perfusion and CFR in patients with chronic heart failure and in healthy volunteers by measuring coronary sinus flow at VEC MR imaging.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTS DISCUSSIONREFERENCES |
|---|
Cine MR Imaging for LV Mass, Volumes, and Function
All images were acquired with a 1.5-T MR imager (Signa; GE Medical Systems, Milwaukee, Wis) with a phased-array chest coil to improve spatial resolution. Four electrocardiographic leads and bellows were attached to the patient for cardiac gating and respiratory monitoring. Double oblique short-axis images of the left ventricle were acquired in 16 successive cardiac cycles during a one-breath-hold k-space–segmented fast cine MR sequence (FASTCARD; GE Medical Systems). Imaging parameters included the following: 17/2.8 (repetition time msec/echo time msec), section thickness of 10 mm, field of view of 36 x 36 cm, and image matrix of 256 x 256. Nine to 11 contiguous short-axis MR images were acquired from the apex to the base to cover the entire left ventricle. With retrospective gating, 12 images were reconstructed at each anatomic location to represent the cardiac cycle. This data reconstruction strategy resulted in a temporal resolution of 78 msec per image.
VEC MR Imaging for Coronary Sinus Flow Measurement
The coronary sinus was localized in the atrioventricular groove on basal short-axis cine MR images. Coronary sinus flow measurements were obtained with a non–breath-hold VEC phase-contrast gradient-echo sequence with k-space segmentation (Cine PC; GE Medical Systems). The imaging plane was placed perpendicular to the coronary sinus, approximately 2 cm proximal to the entrance of the coronary sinus into the right atrium (10). A retrospective cardiac-gating strategy enabled ongoing data acquisition throughout the cardiac cycle, and frames at the end of the cycle were time resolved, as were those obtained at the beginning (11). Imaging parameters included the following: 27/7.9, flip angle of 30°, section thickness of 5 mm, two signals acquired, field of view of 24 x 24 cm, and image matrix of 256 x 256, which resulted in an in-plane resolution of 0.94 x 0.94 mm. Phase wrap was avoided by means of oversampling of twice the prescribed field of view in the phase-encode direction with two signals acquired. The resulting image was then cropped to keep the center 256 imaging lines with the given size of pixels.
A complete set of VEC MR images was acquired in 5 minutes. Data were sorted according to their occurrence during the cardiac cycle and were interpolated into 16 phase and magnitude images (12). Coronary sinus flow measurements were performed at rest with velocity encoding set to 100 cm/sec. Subsequently, dipyridamole (Boehringer Ingelheim, Ridgefield, Conn) (0.56 mg per kilogram of body weight) was infused into an antecubital vein over 4 minutes to induce coronary vasodilation. Velocity encoding was changed to 200 cm/sec, and image acquisition was started 2 minutes after administration of dipyridamole to avoid underestimation of peak stress flow. Heart rate and blood pressure were monitored and recorded during the entire protocol. Serious side effects, such as angina pectoris, dyspnea, or ventricular tachycardia, were documented if they occurred after infusion of dipyridamole. After conclusion of the imaging protocol, the effect of dipyridamole was reversed by injecting 75 mg of aminophylline.
Data Analysis
MR images were transferred via Ethernet to a Macintosh computer (Apple Computers, Cupertino, Calif), and data analysis was performed with a public domain program (NIH Image, version 1.59; U.S. National Institutes of Health; available at rsb.info.nih.gov/nih-image/). All MR measurements were performed independently by two of three observers (G.K.L., N.W., M.Y.). Data are given as the mean values for the two observers, and interobserver variability was calculated.
To evaluate LV mass, the epicardial and endocardial borders were manually traced on end-diastolic images to include the papillary muscles at each anatomic level to encompass the left ventricle. LV mass was calculated by summing the myocardial volume areas and multiplying by the density (1.05 mg/mL) of myocardial tissue (13).
Coronary sinus flow was measured by tracing the contour of the coronary sinus on each magnitude image throughout the cardiac cycle (Figs 1, 2). Systole and diastole were determined on the basis of opening and closing of the aortic and mitral valves, which are shown on the magnitude MR images (Fig 1). Care was taken to closely follow the boundary of the vessel identified by a change in signal intensity on magnitude images as a result of surrounding epicardial fat or myocardium (9). The area of coronary sinus was recorded, and the traced region was transferred to corresponding phase images to measure spatial average flow velocity (Fig 2). Phasic blood flow was calculated as the product of area and spatial average flow velocity. Mean volume flow was derived by means of integration of phasic flow over time. Figure 3 shows representative phase images of the area of the coronary sinus and the velocity of coronary sinus blood flow and volume flow in a volunteer and a patient with chronic heart failure.
|
|
|
|
|
|
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTS DISCUSSIONREFERENCES |
|---|
|
After infusion of dipyridamole, no patients or volunteers experienced any deleterious side effects. The systolic and diastolic blood pressures did not change significantly in either group compared with those at rest. Compared with the rate at rest, the heart rate increased significantly to 98 beats per minute ± 11 in volunteers (P = .001) and to 80 beats per minute ± 11 in patients (P = .001), and the rate-pressure product increased to 11,020 mm Hg/min ± 1,063 in volunteers (P = .001) and to 10,127 mm Hg/min ± 1,818 in patients (P = .001). The rate-pressure product after administration of dipyridamole was not significantly different between the two groups (P = .14).
Coronary Sinus Flow at Rest and after Infusion of Dipyridamole
At rest, LV perfusion was not significantly different between patients with chronic heart failure (0.46 mL/min/g ± 0.19) and volunteers (0.52 mL/min/g ± 0.21) (P = .54). After administration of dipyridamole, LV perfusion was significantly increased to 1.07 mL/min/g ± 0.64 in patients and to 2.19 mL/min/g ± 0.98 in volunteers (P = .01 for both groups vs values at rest) (Fig 4). The increment of LV perfusion after dipyridamole infusion was significantly smaller in patients than that in volunteers (P = .03) (Fig 4). Furthermore, CFR was significantly reduced in patients compared with that in volunteers (2.3 ± 0.9 vs 4.2 ± 1.5, P = .01) (Fig 5). Regression analysis revealed a moderate positive relationship between CFR and LV ejection fraction (Y = 0.84 ± 0.05X, r = 0.54, P = .02, standard error of the estimate = 1.34) (Fig 6). Interobserver variability for measurement of coronary sinus flow was 3% ± 8 at rest and 1% ± 7 after infusion of dipyridamole.
|
|
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTS DISCUSSIONREFERENCES |
|---|
Comparison with Previous Studies
In two previous studies, normal data are reported for LV perfusion in volunteers with VEC MR measurements of coronary sinus flow (10,15). Similar to the current findings, Schwitter et al (10) found LV perfusion in volunteers of 0.53 mL/min/g ± 0.14 at rest, which increased to 2.27 mL/min/g ± 0.78 after infusion of dipyridamole. Kawada et al (15) reported slightly higher values of 0.74 mL/min/g ± 0.23 for LV perfusion in volunteers at rest and 2.14 mL/min/g ± 0.51 after infusion of dipyridamole. The good agreement of perfusion data between the current study and previous studies indicates the robustness of MR imaging for quantifying LV perfusion. Measurements of LV perfusion are slightly lower at MR imaging than those obtained at PET, which ranged between 0.75 and 1.1 mL/min/g at rest in volunteers (16).
Remodeling and LV Perfusion in Patients with Chronic Heart Failure
A number of studies were performed to investigate the development of cardiac remodeling after myocardial infarction (5,6,17). Most previous MR studies have focused on the effect of medical therapy on remodeling in patients with acute or chronic infarction (18,19). To our knowledge, however, little is known about the effect of LV perfusion on LV remodeling. Findings in an experimental study revealed that the myocardial blood flow is reduced after vasodilation in only those animals with LV remodeling and features of chronic heart failure (20). Interestingly, animals without any features of chronic heart failure had normal LV perfusion at rest and after vasodilation (20). Results of a recent clinical study show a positive relationship between preserved CFR and extent of viable myocardium in patients after acute myocardial infarction (4). It is conceivable that a preserved CFR implies survival of viable myocardium, which in turn mitigates progressive remodeling after myocardial infarction.
Data collected in the current study give further insight into the relationship between LV perfusion and cardiac remodeling in patients with chronic heart failure. Patients in the current study had typical features of remodeling, such as increased LV mass, increased end-systolic volume, and decreased ejection fraction. Despite increased myocardial mass, LV perfusion at rest was normal in patients compared with that in volunteers (Fig 4). However, dipyridamole showed a severely reduced CFR in this group of patients, which suggests a profound perfusion abnormality that is most likely related to multiple stenoses in the coronary arteries. Patients with the most severely depressed CFR had the lowest contractile function (Fig 6). The positive correlation between reduced CFR and impairment of LV function found in the current study emphasizes the negative effect of reduced perfusion on myocardial function.
Contributing Factors for Reduced CFR
Besides stenoses of coronary arteries, there may be several other reasons for reduced CFR in patients with chronic heart failure. First, CFR is reduced in patients with chronic heart failure because of increased LV mass. Findings in a previous study show that patients with LV hypertrophy have a higher level of autoregulated blood flow at rest to meet the demand of the increased myocardial mass (2). Therefore, CFR is reduced in patients with LV hypertrophy because the resting flow is already raised closer to the level of maximal achievable flow (2). Second, recurrent microembolization may be responsible for a reduction in CFR and a decrease in contractile function in patients with coronary artery disease (21,22). Results of a study of patients with coronary artery disease revealed that CFR was persistently reduced despite a successfully performed coronary angioplasty, presumably as a result of microembolization into distal segments of the dilated coronary artery (21). Furthermore, findings in an animal study demonstrate a progressive decline in contractile function after experimentally induced microembolization (22). Third, impaired endothelium-dependent vascular dilatation of the precapillary vessels (23,24) or increased neurohormonal sympathetic activity may contribute to reduced CFR in patients with chronic heart failure (25,26). Fourth, results in one study (27) show that the responsiveness of the coronary vascular bed to a vasodilator is curtailed mainly by capillary resistance, which is regulated by changes in capillary dimensions or by derecruitment of capillary vessels. Capillary resistance is affected whenever there is functional or structural damage to the capillary vessels, such as in myocardial infarction, diabetes, or hypertension (28). Variable hemodynamics did not noticeably influence the difference in CFR because the rate-pressure products were identical for the two groups at rest and during administration of dipyridamole.
Limitations
The MR imaging strategy used in the current study does not provide regional assessment of LV perfusion, but first-pass perfusion MR imaging can be used when regional assessment of myocardial blood flow is crucial. Note that the development and progression of coronary artery disease is not limited to individual arteries but is a global myocardial process. Similarly, cardiac remodeling comprises functional and architectural changes of the entire heart. To study the relationship between remodeling and perfusion of the left ventricle, it was important to quantify global instead of regional myocardial perfusion. Measurement of LV perfusion in patients with chronic heart failure as a result of inferior wall infarction may be difficult with the proposed imaging strategy, because a variable part of the posterior and inferior septal wall drains into the coronary sinus just before it empties into the right atrium (16). This variable fraction of LV perfusion could have been missed because measurements were performed approximately 2 cm before the entrance of the coronary sinus into the right atrium.
Note also that the number of patients in the current study was small, and they had only mild to moderate symptoms of chronic heart failure. Differences between the groups may have been more defined if patients with severe symptoms of chronic heart failure had been included.
The volunteers were substantially younger and had a lower body surface area compared with the patients. Young volunteers were studied to make it less likely that any of these subjects had coronary artery disease. After we controlled for possible confounding effects of age and body surface area by means of analysis of covariance, the differences were still present between groups. However, a better comparison would have been achieved with volunteers with matched ages.
Accurate flow measurements are dependent on precise definition of vessel size. The vessel size was defined by outlining the border of the vein on the magnitude image. The decision to include a pixel was based on the brightness of the pixel, which represents a validated approach to analyze the data (9,10). Reliability of this approach is confirmed with the low interobserver variability.
Clinical Implications and Conclusions
In the current study, we propose a noninvasive approach to simultaneously evaluate LV function and global LV perfusion in patients with chronic heart failure that provides further insight into the development of cardiac remodeling. Serial noninvasive measurements of mass, volume, and function, as well as myocardial perfusion at rest and during hyperemic conditions, should improve knowledge about the progression of LV remodeling. Furthermore, this imaging strategy may be useful for studying the effect of new interventional or pharmacologic therapies designed to improve LV perfusion and to ameliorate cardiac remodeling.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
Author contributions: Guarantors of integrity of entire study, G.K.L., M.S., M.B., C.B.H.; study concepts, G.K.L., M.B., C.B.H.; study design, G.K.L., M.S., M.B., C.B.H.; literature research, G.K.L., M.S.; clinical studies, G.K.L., N.W., M.Y., G.P.R., P.A.A., D.C.; data acquisition, G.K.L., N.W., M.Y., G.P.R., P.A.A.; data analysis/interpretation, G.K.L., N.W., M.Y., P.A.A.; statistical analysis, G.K.L., N.W.; manuscript preparation, G.K.L., M.S., C.B.H.; manuscript definition of intellectual content, G.K.L., M.S., M.B., C.B.H.; manuscript editing, G.K.L., M.S., C.B.H.; manuscript revision/review and final version approval, all authors.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTS DISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  N. Beohar, A. K. Erdogan, D. C. Lee, H. N. Sabbah, M. J. Kern, J. Teerlink, R. O. Bonow, and M. Gheorghiade Acute heart failure syndromes and coronary perfusion. J. Am. Coll. Cardiol., July 1, 2008; 52(1): 13 - 16. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. W. Losordo, R. A. Schatz, C. J. White, J. E. Udelson, V. Veereshwarayya, M. Durgin, K. K. Poh, R. Weinstein, M. Kearney, M. Chaudhry, et al. Intramyocardial Transplantation of Autologous CD34+ Stem Cells for Intractable Angina: A Phase I/IIa Double-Blind, Randomized Controlled Trial Circulation, June 26, 2007; 115(25): 3165 - 3172. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| RADIOLOGY | RADIOGRAPHICS | RSNA JOURNALS ONLINE |
) and after infusion of dipyridamole (
). Coronary sinus is depicted with a large number of pixels, which enables close tracing of the vessel throughout the cardiac cycle. Note dilatation of the coronary sinus in the patient. At rest, blood flow velocities show a biphasic pattern, with a first peak during midsystole and a second peak during early diastole. Negative velocities occur at end diastole because of backward flow into the coronary sinus as a result of right atrial contraction. After administration of dipyridamole, the increase in flow velocity is more pronounced in the volunteer, with peak flow velocities at end systole and early diastole. In c, mean coronary sinus flow is represented by the area under the volume flow curve. Note that coronary sinus flow at rest is lower in the volunteer (126 mL/min) than in the patient (173 mL/min), which is related to the smaller LV mass in the volunteer. Normalized LV perfusion per myocardial mass is identical in the volunteer and the patient. After administration of dipyridamole, mean coronary sinus flow increases to 528 mL/min in the volunteer and 545 mL/min in the patient, which represents CFRs of 4.2 and 3.2, respectively.
) and patients with chronic heart failure (