| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
1
Institut für Röntgendiagnostik & Medizinische Klinik,
Universität Würzburg, Josef-Schneider-Straße 2, D-97080
Würzburg, Germany.
2
Department of Cardiovascular Medicine, University of Oxford, John Radcliffe
Hospital, Headley Way, Oxford, OX3 9DU, United Kingdom.
Received August 27, 2001;
accepted after revision October 18, 2001.
Supported by a grant from the "Deutsche Forschungsgemeinschaft,
Sonderforschungsbereich 355: Pathophysiologie der Herzinsuffizienz,"
Part A3.
Abstract
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
MATERIALS AND METHODS. Twelve patients with aortic stenosis were examined with an electrocardiographically triggered two-dimensional tagging sequence at 1.5 T before and 12 months after surgical valve replacement for the evaluation of wall function of the apical, mid ventricular, and basal levels. Eight healthy volunteers in the same age group served as the control group.
RESULTS. Before surgery, all patients showed a significant increase of apical rotation (22.2° ± 5.9° vs 10.3° ± 2.5°, p < 0.0001) and overall left ventricular torsion (25.1° ± 6.6° vs 14.5° ± 3.7°, p < 0.001); basal rotation was not significantly different (-2.9° ± 2.1° vs -4.2° ± 1.9°, p = not significant) compared with the volunteer group. Apical rotation and torsion were negatively correlated with left ventricular mass (r = -0.73, p < 0.01, and r = -0.61, p < 0.05, respectively) and end-diastolic volume (r = -0.73, p < 0.01 and r = -0.64, p < 0.03, respectively). One year after surgery, basal rotation was reduced in the patients with aortic stenosis compared with the patients in the control group (-1.9° ± 1.8°, p < 0.01). In comparison with preoperative values, apical rotation (14.2° ± 3.6°, p < 0.01) also decreased but was still elevated, and this resulted in a normalization of left ventricular torsion (16.1° ± 3.7°, p < 0.01).
CONCLUSION. Surgical valve replacement for aortic stenosis leads to normalization of the left ventricular torsion 1 year after surgery. Pressure overload before surgery is associated with an increase of systolic left ventricular wringing motion, possibly serving as a compensatory mechanism. This mechanism declines with increasing left ventricular hypertrophy and dilatation.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
MR tagging studies showed that in patients with aortic stenosis, systolic rotation is reduced at the base and is increased at the apex, and maximal systolic torsion is also increased [8, 10]. Additionally, aortic stenosis leads to cardiac hypertrophy and impaired global left ventricular function with an improvement after surgical valve replacement, as determined by MR cine studies [12]. However, the response of left ventricular rotation and contraction characteristics after surgical valve replacement is unknown. The aim of our study was therefore to determine the changes in the cardiac systolic rotation and contraction mode in patients before and after surgical valve replacement for aortic stenosis in comparison with healthy volunteers in the same age group. Thus we addressed the question of whether MR tagging provides clinically relevant parameters that characterize alterations of cardiac tissue function in aortic stenosis and, if so, how such parameters change over time when chronic pressure overload is relieved by surgical valve replacement.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Image Acquisition
MR imaging was performed on a 1.5-T scanner (Magnetom Vision; Siemens,
Erlangen, Germany) with 25-mT/m maximum gradient strength and a phased array
body coil. Patients and volunteers were studied in the supine position.
Short-axis cine MR imaging of the entire left ventricle from the base to the
apex was performed for the analysis of cardiac mass and function. A segmented
two-dimensional electrocardiographically triggered fast low-angle shot-pulse
sequence was used during breath-holding (field of view, 240 x 320
mm2; matrix, 126 x 256; TR/TE, 9.9/4.8; flip angle, 30°).
Slice thickness was 8 mm without interslice gap, and the number of slices
acquired ranged from 9 to 11. The acquisition window for one cardiac phase was
80-100 msec, resulting in a temporal resolution of 40-50 msec because of echo
sharing. The number of cardiac phases imaged depended on heart rate;
typically, 16 phases per cardiac cycle were obtained.
Tagging MR imaging was performed in the short heart axis orientation with a slice thickness of 8 mm during a breath-hold. Three short axis planes were obtained; slice positions were taken from the cine MR imaging data set. The basal slice was the first to show circumferential myocardium at both diastole and systole. The apical slice was the last to show the intracavity blood pool in all phases over the cardiac cycle. The mid ventricular slice was positioned half-way between the basal slice and the apical slice. A segmented two-dimensional electrocardiographically triggered fast low-angle shot-pulse sequence (field of view, 240 x 320 mm2; matrix, 216 x 256; 9.0/4.0; flip angle, 15°) was used in the cine mode. A rectangular grid with a spacing of 8 mm was applied. The acquisition window for one cardiac phase was 70-90 msec, resulting in a temporal resolution of 35-45 msec that was achieved because of echo sharing. The number of cardiac phases imaged depended on the heart rate.
Image Analysis
Tagged images were processed using the semiautomatic ARGUS software (V34B;
Siemens, Erlangen, Germany). The endo- and epicardial contours were defined in
the end-diastolic frame and the software then detected the tag grid using an
affine plus anisotropic radial scaling transform algorithm. The grid was
adapted to each of the acquired images from the end-diastolic to the
endsystolic frame. If necessary, manual correction was performed by moving,
adding, or deleting tag intersections. Evaluation time for tracking the tag
intersections of one slice was about 30 min. The coordinate data were exported
into a text file that was analyzed by a PC-based dedicated customwritten
software (obtained by T. Johnson). First, the short-axis cross section was
divided into two segments, each for anterior, lateral, posterior, and septal
wall; adaptation of these segments to the image was possible by entering a
reference sector placed at the anterior border of the free right ventricular
wall. Second, the centroid was calculated for each image, and the movement of
the tag intersections was further analyzed in relation to the present actual
centroid. For the determination of radial and angular displacement, the
program converted the two-dimensional linear coordinates into the radius and
angle of each tag intersection relative to the centroid. Rotation and
contraction were calculated by subtracting the end-systolic and end-diastolic
angle or radius, respectively, of each of the tag intersections in one wall
segment. Circumferential shortening was determined by tracking the tag
intersections at the boundaries of the wall segments and calculating the
relative difference between end-diastolic and end-systolic circumference of
the inner and outer myocardium. The output of the program showed a diagram of
the tags' movement and the multiple plots of the circumferential shortening
and of the angular (rotation) and radial (contraction) displacement. The data
were expressed as average values for the entire myocardium for degrees of
rotation and distance of contraction or relative circumferential shortening at
the basal, mid ventricular, or apical levels, respectively, between
end-diastole and end-systole. Torsion was calculated as the maximal rotation
of the apex compared with the base at the same time point. All these
calculations were performed within 1 min without further need for user
interaction.
Left ventricular function and mass were analyzed using the ARGUS software version VB31B (Siemens). After the determination of the end-diastolic and end-systolic frame on the first basal slice to show circumferential myocardium at both diastole and systole, the endocardial and epicardial contours were traced manually by two investigators. The papillary muscles were included in the measured ventricular volume. To account for body height and mass, indexes were calculated related to 1 m2 of body surface area for volumes and masses. Parameters of global function of the left ventricle were end-diastolic and end-systolic volume index and ejection fraction. The ventricular mass index was obtained by the multiplication of the mean wall volume of the end-diastolic and the end-systolic frames and the specific weight of cardiac muscle (1.05 g/mL). All volumes were calculated automatically by summing the areas in the entire series of short-saxis cine images.
All data are presented as means ± SD. For statistical analysis, the Mann-Whitney test was used to identify differences between volunteers and patients. Correlation coefficients were used to study relations between tagging data and parameters of cardiac mass and functions provided by cine MR imaging or invasive measurements, respectively. Wilcoxon's matched pairs test was used to evaluate serial changes after surgical valve replacement. A p value less than 0.05 was considered statistically significant.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Correlation analysis of functional and morphologic parameters obtained by cine MR imaging with the tagging data revealed a significant negative correlation of apical rotation and torsion with the mass index and end-diastolic volume index for patients before surgical valve replacement. Basal and mid ventricular circumferential shortening correlated negatively with the indexes for mass and end-diastolic and end-systolic volumes and positively with the ejection fraction before and after surgical valve replacement, except for basal circumferential shortening and the end-diastolic volume index. Details are shown in Table 2. In healthy volunteers, there was a significant correlation of apical circumferential shortening with the ejection fraction (r = 0.74, p than 0.04) and a negative correlation of basal circumferential shortening with left ventricular mass index (r = -0.94, p = 0.001).
|
Analysis of systolic rotation and contraction for all patients before surgery and 12 months after surgery in comparison with healthy volunteers (Fig. 1A,1B,1C) is shown in Table 3. At study entry, the patients with aortic stenosis had significantly higher values for apical rotation (than p < 0.001) and torsion (than p < 0.01). Lower values were found for basal and apical contraction (than p < 0.01) and for basal and apical circumferential shortening (p < 0.02 and p < 0.03, respectively). Twelve months after surgical valve replacement, there was a significant decrease of apical rotation (p < 0.01) and an increase of basal contraction (p < 0.03) without normalization, whereas basal circumferential shortening increased (p < 0.04) and torsion decreased (p < 0.01) with no remaining significant difference to the healthy volunteers (Fig. 2A,2B,2C,2D). Basal rotation (p < 0.03) was significantly reduced only at follow-up.
|
|
|
|
|
|
|
|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
In patients with aortic stenosis, a reduction of systolic rotation of the base to -2.4° ± 2.0° and an increase of apical rotation to 12.0° ± 6.0° and maximal torsion to 14° ± 5° were described [9, 10]. With our patient group, similar observations were made for the basal level, but values for apical rotation and torsion were remarkably higher than previously reported. Because the comparison of the clinical patient characteristics shows no differences in the stage of disease, as determined by invasive measurements, these discrepancies again might be caused by the different definition of the position of the apical slice. An explanation for the characteristic wringing motion of the left ventricle is the helical muscle fiber orientation that generates high intracavitary pressure with minimal shortening and thus results in minimal energy demand of the muscle fibers. The reduction of basal rotation in aortic stenosis might be due to the stiffening of the valvular plane and myocardial hypertrophy, whereas the increase in apical rotation and torsion is possibly a compensatory mechanism to increase intracavitary pressure [10]. There was no correlation of tagging data with invasive measurements at study entry. This finding suggests that the extent of elevated intracavitary pressure is not the only factor that influences the degree of the compensatory mechanism of increased myocardial rotation and torsion and the decrease of contraction. Analysis of left ventricular mass and function have shown that the left ventricular function was only slightly reduced (ejection fraction, 57% ± 15% before surgery), although contractility was considerably decreased. Left ventricular mass was increased to 125 ± 27 g/m2 before surgery in what can be considered as an adaptation to increased wall stress [13]. However, in our patients, increasing left ventricular hypertrophy and increasing dilatation were accompanied by a decrease of apical rotation and torsion only at study entry and of circumferential shortening at both study entry and follow-up, indicating depressed myocardial contractility. It remains unclear whether an increase of left ventricular mass and volume leads to a decrease of systolic left ventricular wringing motion or whether a smaller degree of this compensatory mechanism results in left ventricular hypertrophy and dilatation. The lack of improvement of basal rotation after surgical valve replacement can be due to the still present stiffening of the valvular plane and to the changes in fiber architecture caused by the implantation of the valve prosthesis itself. The reduction of apical rotation and torsion and the normalization of circumferential shortening, on the other hand, might be caused by the reduced pressure load of the left ventricular. Although apical rotation is still slightly elevated compared with the control subjects, the combination of reduced basal contraction results in values for myocardial torsion after surgical valve replacement that are comparable with healthy volunteers without considerable differences.
Correlation of the tagging data to the changes of morphologic and functional left ventricular parameters before and after surgical valve replacement revealed that the increase of the ejection fraction and the decrease of the mass index and the end-diastolic and end-systolic volume index is correlated with an increase of basal and mid ventricular circumferential shortening. There was no correlation for torsion and rotation and only a correlation of ejection fraction and mid ventricular contraction. Thus, circumferential shortening seems to be the most representative parameter for global cardiac function, at least in patients with aortic stenosis. However, in volunteers there was only a correlation for ejection fraction and mass index with apical and basal circumferential shortening. These results underline that global cardiac function results from a combination of contraction, circumferential shortening, rotation, and torsion.
Limitations and Clinical Implications
Because the applied spatial modulation of the magnetization technique has a
limited tag contrast, we could not evaluate the diastolic cardiac untwisting
that is described to be delayed and prolonged in aortic stenosis
[9,
10]. The use of a
complementary spatial modulation of the magnetization technique with increased
tag contrast and persistence should solve this problem in future studies
[14]. Furthermore, no slice
correction algorithm was used for the translational motion of the heart that
might lead to imaging of different regions in diastole and systole. However,
because the changes in cardiac function due to aortic stenosis involve the
entire left ventricle, this translational motion should not have a
considerable impact on the evaluation of cardiac rotation. One important
feature of the custom-written program is that the centroid is calculated for
each image. Thus the movement of the tag intersections is analyzed in relation
to the present actual centroid and no artifacts can occur because of the
movement of the heart during contraction in the imaging plane.
Despite the previously mentioned technical limitations, the tagging technique allowed the assessment of the changes in left ventricular rotation in patients with aortic stenosis before and after surgery in comparison with healthy age-matched volunteers. These observations are exclusively related to the aortic stenosis because substantial aortic incompetence or coronary artery disease were ruled out by invasive left heart catheterization and coronary angiography. There were no correlations between invasive pressure measurements and tagging data, but there were negative correlations of rotation and contractility with left ventricular hypertrophy and dilatation. Therefore, morphologic parameters have a higher impact on myocardial wall function than do left ventricular pressure determinants. Aortic stenosis is associated with an increase of systolic left ventricular wringing motion, possibly serving as a compensatory mechanism for pressure overload. This compensatory mechanism declines with increasing left ventricular hypertrophy and dilatation. Surgical valve replacement leads to a reduction of basal rotation, whereas left ventricular torsion and circumferential shortening is normalized, indicating the reduction of pressure overload after surgery. Therefore, increased rotation and torsion might be an additional relevant parameter for the stage of disease in patients with only slightly depressed left ventricular function. Thus this might answer the still open question concerning the optimal time for surgery in aortic stenosis.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  I. K. Russel, M. J.W. Gotte, J. G. Bronzwaer, P. Knaapen, W. J. Paulus, and A. C. van Rossum Left ventricular torsion an expanding role in the analysis of myocardial dysfunction. J. Am. Coll. Cardiol. Img., May 1, 2009; 2(5): 648 - 655. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 W.-J. Kim, B. H. Lee, Y. J. Kim, J. H. Kang, Y. J. Jung, J.-M. Song, D.-H. Kang, and J.-K. Song Apical Rotation Assessed by Speckle-Tracking Echocardiography as an Index of Global Left Ventricular Contractility Circ Cardiovasc Imaging, March 1, 2009; 2(2): 123 - 131. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 T. C. Gillebert and N. R. Van de Veire About left ventricular torsion, sex differences, shear strain, and diastolic heart failure Eur. Heart J., May 2, 2008; 29(10): 1215 - 1217. [Full Text] [PDF]
|
|
|
|
|
|
W. G. van Dockum, J. P.A. Kuijer, M. J.W. Gotte, F. J. ten Cate, J. M. ten Berg, A. M. Beek, J. W.R. Twisk, J. T. Marcus, C. A. Visser, and A. C. van Rossum Septal ablation in hypertrophic obstructive cardiomyopathy improves systolic myocardial function in the lateral (free) wall: a follow-up study using CMR tissue tagging and 3D strain analysis Eur. Heart J., December 1, 2006; 27(23): 2833 - 2839. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 T. Helle-Valle, J. Crosby, T. Edvardsen, E. Lyseggen, B. H. Amundsen, H.-J. Smith, B. D. Rosen, J. A.C. Lima, H. Torp, H. Ihlen, et al. New Noninvasive Method for Assessment of Left Ventricular Rotation: Speckle Tracking Echocardiography Circulation, November 15, 2005; 112(20): 3149 - 3156. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M.W. Ashford Jr, W. Liu, S.J. Lin, P. Abraszewski, S.D. Caruthers, A.M. Connolly, X. Yu, and S.A. Wickline Occult Cardiac Contractile Dysfunction in Dystrophin-Deficient Children Revealed by Cardiac Magnetic Resonance Strain Imaging Circulation, October 18, 2005; 112(16): 2462 - 2467. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Vogel, G. Derrick, P. A. White, S. Cullen, H. Aichner, J. Deanfield, and A. N. Redington Systemic ventricular function in patients with transposition of the great arteries after atrial repair: a tissue Doppler and conductance catheter study J. Am. Coll. Cardiol., January 7, 2004; 43(1): 100 - 106. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
) with
recalculation of centroid (+) for each image. Analysis of rotation and
contraction calculated by subtraction of angle or radius, respectively, at
end-diastole and end-systole reveals counterclockwise rotation of 6.9° and
radial contraction of 5.1 mm. Mean circumferential shortening calculated from
percentual difference between end-diastolic and end-systolic circumference of
four segments (boundaries marked by [UNK]) is 25.8%.
