| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Cardiac Imaging |
1 From the Center for Integrated Non-Invasive Cardiovascular Imaging of the Departments of Radiology (Section of Cardiovascular Imaging) (B.B.C., R.M.S., A.E.S., J.A.W., A.G.L., R.D.W.), Cardiovascular Medicine (R.C.S., J.B.Y., R.D.W.), and Thoracic and Cardiovascular Surgery (N.G.S., P.M.M., R.D.W.), and the Department of Biostatistics and Epidemiology (M.L.L.), Cleveland Clinic Foundation, Desk Hb6, 9500 Euclid Ave, Cleveland, OH 44195. Received August 26, 2005; revision requested October 25; revision received December 2; accepted January 5, 2006; final version accepted February 10. B.B.C. supported by the RSNA Research and Education Foundation. Address correspondence to R.D.W. (e-mail: richard.white{at}jax.ufl.edu).
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONADVANCES IN KNOWLEDGEReferences |
|---|
Materials and Methods: The study was performed with institutional review board approval, and a waiver of individual informed consent was obtained. The study was HIPAA compliant. Patients (83 men, 22 women; mean age, 61 years ± 9 [standard deviation]) were evaluated with MR imaging before and after SVR as follows: pre-SVR examination (n = 105; 25 days ± 39 before SVR; median, 7 days; range, 1–189 days), early post-SVR examination (n = 95, 7 days ± 3 after SVR), and late post-SVR (n = 35, 313 days ± 158 after SVR). Cine MR imaging allowed calculation of ejection fraction and rate-corrected velocity of circumferential fiber shortening (VcfC) for global LV functional evaluation, whereas tagged MR imaging (spatial modulation of magnetization with harmonic phase analysis) permitted assessment of regional circumferential strain (EC) with coronary distribution. VcfC and EC were computed at both LV base- and mid-LV short-axis levels remote from the site of anteroapical SVR.
Results: Prior to SVR, LV dilatation and diminished global and regional LV function were observed. At early post-SVR examination, VcfC had improved significantly but EC showed a worsening trend overall, although only EC of the right coronary artery at the mid-LV level worsened significantly. At late post-SVR examination, VcfC values were improved when compared with pre-SVR values, although EC showed no statistically significant improvement. When compared with that at early post-SVR examination, however, EC showed significant improvement in two segments: left anterior descending artery and right coronary artery at mid-LV level.
Conclusion: Although volume-based indexes of global LV function improve significantly after SVR, regional LV function did not improve significantly; there was evidence of continued LV remodeling after SVR.
© RSNA, 2006
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONADVANCES IN KNOWLEDGEReferences |
|---|
Surgical ventricular restoration (SVR) has been promoted as a therapeutic approach to large post-MI dysfunctional zones of the left ventricle, with or without aneurysms; this approach addresses the relationship between LV shape and function (6). SVR is performed to attenuate the LV remodeling process by reducing LV volume and, presumably, restoring geometry and fiber orientation of the left ventricle. Often combined with more standard surgical management of chronic ischemic heart disease (eg, coronary revascularization, mitral valve repair for mitral regurgitation, or both), SVR is performed to improve patient survival and quality of life (7).
There remains a need to understand better the beneficial effects of SVR on LV myocardial function. Although magnetic resonance (MR) imaging is recognized as a valuable tool for evaluating LV dysfunction in the setting of chronic ischemic heart disease (8), few studies have assessed the use of MR imaging in patients before and after surgical correction of large post-MI dysfunctional zones, with or without aneurysms, of the left ventricle (9).
Thus, the purpose of this study was to use dynamic MR imaging to retrospectively evaluate the changes in global and regional LV function after SVR in patients with chronic ischemic heart disease who had large nonaneurysmal or aneurysmal post-MI zones.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONADVANCES IN KNOWLEDGEReferences |
|---|
50% scarring were not included; such segments are not expected to demonstrate improved function after revascularization (10). Furthermore, patients with MI-related zones (with or without aneurysms or pseudoaneurysms) of the lateral or inferior walls were also excluded. All 105 patients underwent both a pre-SVR MR imaging examination (25 days ± 39 before SVR; median, 7 days; range, 1–189 days) and either one or two post-SVR MR imaging examinations. None of the patients had aortic regurgitation as a confounding factor in functional assessments. Post-SVR examinations were divided according to time after pre-SVR examination: early (<25 days after SVR; n = 95; mean, 7 days ± 3) and late (>120 days after SVR; n = 35; mean, 313 days ± 158); no follow-up examinations were performed during the interval of 25–120 days. Seventy patients underwent only early post-SVR examination, 10 underwent only late post-SVR examination, and 25 underwent both early and late post-SVR examination.
Surgical Technique
SVR was performed by using a modification of the Dor technique (11). This technique involved the excision of the scarred anteroapical post-MI zones of the left ventricle followed by approximation of the free edges with circumferential endocardial and epicardial sutures, and a patch if necessary, to restore LV configuration. SVR was performed alone (n = 6), combined with coronary artery bypass grafting (n = 46), combined with mitral valve repair (n = 11), or combined with both coronary artery bypass grafting and mitral valve repair (n = 42).
MR Imaging
MR imaging was performed with a commercially available 1.5-T imager (Magnetom Sonata; Siemens Medical Solutions, Erlangen, Germany) by using torso coils and a standard electrocardiographic gating system.
Cine MR image loops (12) were acquired by using either an electrocardiographic-triggered spoiled gradient-echo pulse sequence (n = 40 at pre-SVR examination, n = 31 at early post-SVR examination, n = 8 at late post-SVR examination) (repetition time msec/echo time msec, 4–6/9–13; flip angle, 30°; section thickness, 6–10 mm; field of view, 300–360 mm at 75%–100% rectangular matrix starting at 256 x 256), an electrocardiographic-triggered balanced steady-state free precession pulse sequence (n = 60 at pre-SVR examination, n = 59 at early post-SVR examination, n = 26 at late post-SVR examination) (1.5–1.65/3.0–3.3; flip angle, 57–60°; section thickness, 8–10 mm; field of view, 263–360 x 300–360 mm; initial matrix, 256 x 256), or a retrospective electrocardiographic-triggered balanced steady-state free precession sequence (n = 5 at pre-SVR examination, n = 5 at early post-SVR examination, n = 1 at late post-SVR examination) (1.6/3.5; flip angle, 70°; section thickness, 8–10 mm; field of view, 263–360 x 300–360 mm; initial matrix, 256 x 256). Temporal resolution of cine image loops varied from 30 to 50 msec. One signal average was used in patients capable of breath holding (10–12 seconds); otherwise three averages were used during free breathing. Dynamic MR image series were acquired in the LV short-axis plane from the mitral valve to the LV apex. Select image loops were also acquired in two-chamber, three-chamber, and four-chamber orientations.
Dynamic tagged MR image loops (12) were acquired in each patient by using a retrospective electrocardiographic-gated, segmented k-space, grid-tagged gradient-echo sequence (spatial modulation of magnetization) with 8-mm tag spacing (4/9; flip angle, 15°; section thickness, 10 mm; field of view, 244–350 x 300–380 mm; initial matrix, 256 x 256; temporal resolution, 40–55 msec) at select short-axis and long-axis locations to match the cine image loops. This sequence was performed with (one average) or without breath holding (three averages).
Dynamic phase-contrast MR imaging (12) was used to measure flow in the ascending aorta and was performed with an electrocardiographic-gated velocity-encoded spoiled gradient-echo pulse sequence (5/24; section thickness, 5–8 mm; flip angle, 30°; field of view, 260–300 x 280–350 mm; initial matrix, 256 x 256). Images were acquired perpendicular to the mid-ascending aorta. To avoid aliasing artifacts, velocity encoding was set at 150 cm/sec. In each patient, 30 frames were acquired per cardiac cycle during free breathing by using three averages.
Last, although not used for LV functional assessment, delayed-enhancement MR imaging (8) (4/8) was also performed to help delineate the anteroapical scarring for eventual surgical excision at SVR (Fig 1).
|
The mid-ascending aorta was manually delineated (B.B.C., A.E.S., R.D.W.) on dynamic phase-contrast MR images in a plane perpendicular to blood flow to calculate net aortic flow during the cardiac cycle (effective stroke volume, SVeff), as well as effective ejection fraction (EFeff) (EFeff = SVeff/end-diastolic volume) and mitral regurgitation fraction (mitral regurgitation [in percent] = SVtot – SVeff/SVtot).
Rate-corrected velocity of circumferential fiber shortening (Vcfc), an ejection phase index sensitive to contractile state that is independent of preload and incorporates afterload, is inversely related to end-systolic LV wall stress in a linear fashion (13). VcfC was computed at the basal and mid-left ventricle as
 |
D (in percent) is fractional shortening of LV short-axis diameter measured at mid-wall, LVET (milliseconds) is ejection time between end diastole and end systole (with closed mitral and aortic valves at both points), and RR (milliseconds) is the R-R interval. To account for the surgical changes in LV shape between examination time points and to reduce the influence of the surgical geometric changes on the results, Vcfc was quantified only at representative LV base and mid-LV short-axis levels remote from the site of surgery. The mid-LV level was defined as the short-axis level where papillary muscle insertion points were depicted throughout the cardiac cycle; the LV base level was defined as halfway between the mid-LV level and the mitral annulus.
Quantitative Image Analysis: Regional LV Function
By using dynamic tagged MR images (14), LV mid-wall circumferential strain (Ec) (15) was computed with harmonic phase analysis (HARP; Diagnosoft, Palo Alto, Calif), which uses isolated spectral peaks in the frequency domain representation of the tagged images (16). By manually delineating the LV endocardium and epicardium at end systole (B.B.C., 1 year of experience in cardiac imaging, under the supervision of R.M.S., 9 years of experience) on short-axis tagged series, the LV mid-wall was specified and then located automatically in all other frames (4). Ec was calculated based on average deformation of mid-wall points. Subsequently, strain results at each level were subdivided into three 120° segments (beginning in the mid-interventricular septum, halfway between the right ventricular insertion points) to approximate the coronary artery distributions (left anterior descending artery, right coronary artery, left circumflex artery) (Fig 2).
|
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONADVANCES IN KNOWLEDGEReferences |
|---|
|
|
|
|
Comparison between Pre-SVR and Early Post-SVR Findings
At early post-SVR examination (Fig 6), the expected global volumetric changes after SVR were noted; there were significant reductions in both end-diastolic volume and end-systolic volume (P < .001 for each) (Table). LV dimensions, however, remained abnormally elevated (17). As anticipated, measures of mitral regurgitation were significantly reduced (P < .001 for each) because of improvement in ischemic regurgitation, frequent mitral valve repair, or both.
|
.017 for each). Both EFtot and EFeff, however, showed significant improvement (P < .001 for each). In addition, VcfC increased significantly at both the LV base and mid-LV level at early post-SVR examination (P < .001 for each) (Fig 3). Overall EC showed a worsening trend (ie, became less negative) at both the LV base and mid-LV level (Fig 4), as did segmental EC in four of six LV segments (all but EC at the left anterior descending artery and EC at the right coronary artery at the base) (Fig 5).
Comparison between Pre-SVR and Late Post-SVR Findings
The effects of SVR are demonstrated in a patient with a nonaneurysmal large anteroapical post-MI zone (Fig 7); images of this patient show both the global changes in LV geometry caused by undergoing SVR and regional improvement in intramyocardial deformation observed in some patients. Statistically significant reductions in LV volume (ie, end-diastolic volume, end-systolic volume) between the pre-SVR examination and the early post-SVR examination were maintained at the late post-SVR examination (P < .001 for each) (Table).
|
.017) when compared with pre-SVR levels. Furthermore, no significant change was observed in either measure of mitral regurgitation (P .017). In contrast, EFtot and EFeff were improved significantly when compared with pre-SVR values (P < .001) (Table). VcfC did not change significantly at the LV base level (P .017). VcfC, however, was increased significantly at the mid-LV level (P = .013) (Fig 3), which reflected the significantly decreased mid-wall LV dimension at end-systole (P = .005); there was no statistically significant change in end-diastolic dimension (P = .017) or R-R interval (P = .16).
There was no significant change in overall EC either at the LV base or mid-LV level between the pre-SVR and late post-SVR examinations (Fig 4), although a trend toward improved strain was observed at both levels (P .017 for each). Similarly, no changes in segmental EC between pre-SVR and late post-SVR examinations were statistically significant (P .017), although there were similar trends toward improvement in all segments (Fig 5).
Comparison between Early Post-SVR and Late Post-SVR Findings
At late post-SVR examination, end-diastolic volume increased significantly when compared with that at early post-SVR examination and remained significantly less than that at pre-SVR examination (P = .005); end-systolic volume did not increase with the same pattern (P .017) (Table). The interval increase in end-diastolic volume alone was associated with significant increases in SVtot (P = .001) and SVeff (P = .005), although there was no significant change in EFtot, EFeff, cardiac output, or cardiac index (P .017). In addition, there was no significant change in VcfC at either the LV base or mid-LV level (P .017) (Fig 3). Measures of mitral regurgitation showed a worsening trend back to pre-SVR levels (P .017).
Overall EC improved significantly at the mid-LV level (P = .002) between the early and late post-SVR examinations; there was also a trend toward improvement at the LV base level (P .017) (Fig 4). Segmental EC also showed a trend toward improvement in all regions at both levels, however, only two changes were statistically significant (Fig 5): EC of the left anterior descending artery (P = .004) and EC of the right coronary artery (P = .015) at the mid-LV level.
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONADVANCES IN KNOWLEDGEReferences |
|---|
In our study of changes in global and regional LV function after SVR for treatment of large post-MI dysfunctional anteroapical scar zones with or without true aneurysms, statistically significant improvements in measures of overall LV function (ie, EFtot, EFeff, and VcfC) were seen in the early post-SVR period and were maintained into the late post-SVR period. These results are consistent with those from prior studies (9,19).
VcfC is a sensitive measure of contractile state that is independent of preload but is highly dependent on afterload (13,20). VcfC has been shown to be more sensitive to impaired contractility than has EF (21). Thus, the sustained improvement in VcfC after SVR found in our study appears to reflect reduced LV wall stress (ie, afterload) as a consequence of the surgically restored ventricular shape, improved myocardial function, or both due to concurrent coronary artery bypass grafting.
Taniguchi et al (22) observed that mean velocity of circumferential shortening in noninfarcted myocardial segments increased after linear LV aneurysmectomy. Additionally, these investigators found a significant correlation between the postsurgical increase in velocity of circumferential shortening and the postsurgical decrease in regional end-systolic LV wall stress. This, they concluded, provided evidence that an improvement of systolic performance in the noninfarcted contractile segments is largely because of a reduction in regional afterload.
A statistically significant increase in end-diastolic volume was seen in our study between the early and late post-SVR examinations, which suggests recurrent LV remodeling. This trend has also been noted in other studies (23,24) and appears to be more problematic in patients with a larger preoperative LV end-systolic volume index (24). In addition, the trend of worsening mitral regurgitation values between the early and late post-SVR examinations in our study may be a further indication of the remodeling process. Other researchers have noted the association of recurrent mitral regurgitation with increasing end-diastolic volume over time after SVR (19).
At early post-SVR examination, EC demonstrated a worsening trend overall when compared with that at pre-SVR examination; one segment deteriorated significantly (EC of the right coronary at the mid-LV level). These results are likely a consequence of the direct trauma of the surgical procedures themselves (SVR, coronary artery bypass grafting, mitral valve repair), the induced systemic inflammatory state associated with cardiopulmonary bypass, the known myocardial stunning or ischemic injury associated with cardiac surgery, or a combination of any of these factors (25–28).
Although statistically significant improvement was found between the early and late post-SVR examinations of EC of the left anterior descending artery and EC of the right coronary artery at the mid-LV level, these changes likely reflect improvement in the depressed function seen at early post-SVR examination. No significant improvement was noted between the pre-SVR and late post-SVR examination. Improvement trends, however, were seen throughout the LV basal and mid-LV segments. In contrast, Kramer et al (9) used dynamic MR imaging to evaluate regional function in patients after linear LV post-MI aneurysm repair but found no significant change in the principal shortening strain (analogous to circumferential strain) in any myocardial regions at 6 weeks after surgery.
In our study, the general trend of improved regional EC at late post-SVR examination is likely of clinical importance. Patients with chronic ischemic heart disease with dilated cardiomyopathy treated with traditional surgical and medical management have a poor prognosis (2-year mortality of 35%–50%) (29,30). Therefore, the trend toward improvements in global and regional EC supports the reported clinical improvements that lead to improved New York Heart Association–functional class, reduced admissions for congestive heart failure, and improved long-term survival (6,31). At the very least this would support a slowing down or attenuation of the progressive LV remodeling expected in these patients. Kokaji et al (24) concluded that although it may not be possible to eliminate the vicious cycle of LV remodeling, progression of LV remodeling can at least be slowed by using the Dor procedure.
It has also been proposed that SVR might improve LV torsion. As the enlarging left ventricle changes from elliptical to spherical, normal systolic torsion is reduced (6,32). SVR changes the spherical LV shape characteristic of congestive heart failure to a more elliptical shape. Theoretically, this reduces wall stress and produces a more helical fiber orientation, which results in increased deformation (ie, torsion) and enhanced thickening of remote viable muscle (33).
Our study had limitations. This study was a retrospective analysis of clinically acquired data. Only three segments per short-axis level were used per level, instead of the six recommended by the American Heart Association. This was done to account for the thinned LV walls in these patients by increasing the number of material points available for strain analysis per segment in order to reduce noise in the EC measurements. Only two short-axis levels were analyzed per patient; however, the surgical procedure precluded finding matching apical levels before and after SVR.
In some patients, pre-SVR cine MR images were acquired by using a gradient-echo pulse sequence, whereas post-SVR cine MR images were acquired by using a balanced steady-state free precession sequence (one [1%] of 95 patients for pre-SVR to early post-SVR comparisons, five [14%] of 35 patients for pre-SVR to late post-SVR comparisons). Because the gradient-echo sequence results in significant underestimation of LV volumes when compared with that of the balanced steady-state free precession sequence (17), use of different acquisition techniques could have resulted in underestimation of changes in the volume-based parameters (including VcfC) between examination time points.
Our study included only patients who underwent SVR with a modified Dor technique, which may not be applicable to other SVR techniques, although our results (eg, VcfC and EC) were comparable to those found in patients studied after linear aneurysm repair (9,22). Furthermore, the follow-up time was short. For instance, the myocardium may take up to 1 year to respond to coronary revascularization (34,35), which the majority of the patients in our study underwent concurrently with SVR. Last, although the overall sample population is large, some of the paired comparisons between examinations at different time points were based on relatively small sample sizes. In particular, only 35 patients underwent late post-SVR examination, which could have affected our results.
In conclusion, we have demonstrated that although volume-based indexes of global LV function improve as a result of SVR, regional LV function (as assessed with circumferential strain) did not improve significantly; there was evidence of continued LV remodeling after SVR. SVR creates a more physiologic LV shape, but the underlying myocardium still shows signs of chronic ischemic heart disease.
|
ADVANCES IN KNOWLEDGE |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONADVANCES IN KNOWLEDGEReferences |
|---|
|
FOOTNOTES |
|---|
Abbreviations: Ec = circumferential strain • EF = ejection fraction • LV = left ventricular • MI = myocardial infarction • SV = stroke volume • SVR = surgical ventricular restoration • VcfC = velocity of circumferential fiber shortening
2 Current address: Department of Cardiothoracic Surgery, Northwestern University Feinberg School of Medicine, Chicago, Ill 
Authors stated no financial relationship to disclose.
Author contributions: Guarantors of integrity of entire study, B.B.C., R.M.S., R.D.W.; study concepts/study design or data acquisition or data analysis/interpretation, all authors; manuscript drafting or manuscript revision for important intellectual content, all authors; approval of final version of submitted manuscript, all authors; literature research, B.B.C., R.M.S., R.D.W.; clinical studies, B.B.C., R.M.S., A.E.S., N.G.S., P.M.M., R.C.S., J.B.Y., J.A.W., A.G.L., R.D.W.; statistical analysis, B.B.C., R.M.S., M.L.L.; and manuscript editing, all authors
|
References |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONADVANCES IN KNOWLEDGEReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  P. P. Agarwal, P. N. Cascade, and F. Pagani Novel Treatment Options for Chronic Heart Failure: A Radiologist's Perspective Am. J. Roentgenol., July 1, 2009; 193(1): W14 - W24. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K. Takeda, G. Matsumiya, H. Matsue, S. Hamada, M. Sakaki, T. Sakaguchi, T. Fujita, and Y. Sawa Use of quantitative analysis of remote myocardial fibrosis with delayed-enhancement magnetic resonance imaging to predict outcomes after surgical ventricular restoration for ischemic cardiomyopathy. J. Thorac. Cardiovasc. Surg., December 1, 2008; 136(6): 1514 - 1521. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 Z.-Z. Song, R. M. Setser, B. B. Carmichael, and R. D. White Surgical Ventricular Restoration and Left Ventricular Function Radiology, February 1, 2008; 246(2): 652 - 653. [Full Text] [PDF]
|
|
|
|
 |
|
 A. S. Blom, J. J. Pilla, J. Arkles, L. Dougherty, L. P. Ryan, J. H. Gorman III, M. A. Acker, and R. C. Gorman Ventricular Restraint Prevents Infarct Expansion and Improves Borderzone Function After Myocardial Infarction: A Study Using Magnetic Resonance Imaging, Three-Dimensional Surface Modeling, and Myocardial Tagging Ann. Thorac. Surg., December 1, 2007; 84(6): 2004 - 2010. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
G. D. Buckberg Congestive heart failure: Treat the disease, not the symptom Return to normalcy/Part II-The experimental approach. J. Thorac. Cardiovasc. Surg., October 1, 2007; 134(4): 844 - 849. [Full Text] [PDF]
|
|
|
|
|
|
R. M. Setser, N. G. Smedira, M. L. Lieber, E. D. Sabo, and R. D. White Left ventricular torsional mechanics after left ventricular reconstruction surgery for ischemic cardiomyopathy. J. Thorac. Cardiovasc. Surg., October 1, 2007; 134(4): 888 - 896. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. J. Gibbons, P. A. Araoz, and E. E. Williamson The Year in Cardiac Imaging J. Am. Coll. Cardiol., September 4, 2007; 50(10): 988 - 1003. [Full Text] [PDF]
|
|
|
|
 |
|
 A. F.L. Schinkel, D. Poldermans, A. Elhendy, and J. J. Bax Assessment of Myocardial Viability in Patients with Heart Failure J. Nucl. Med., July 1, 2007; 48(7): 1135 - 1146. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| RADIOLOGY | RADIOGRAPHICS | RSNA JOURNALS ONLINE |
= Statistical significance of changes between time points (P < .05 vs pre-SVR examination).
= Statistical significance of changes between time points (P < .05 for early post-SVR examination vs late post-SVR examination).
