|
|
||||||||
Ann Thorac Surg 2007;84:1243-1249
© 2007 The Society of Thoracic Surgeons
a Harrison Department of Surgical Research, University of Pennsylvania School of Medicine, Philadelphia, Pennsylvania
b Department of Medicine, University of Pennsylvania School of Medicine, Philadelphia, Pennsylvania
Accepted for publication May 4, 2007.
* Address correspondence to Dr Gorman, Department of Surgery, 313 Stemmler Hall 36th and Hamilton Walk, University of Pennsylvania, Philadelphia, PA 19104-4283 (Email: gormanr{at}uphs.upenn.edu).
| Abstract |
|---|
|
|
|---|
Methods: A well-characterized ovine model of ischemic mitral regurgitation was used. Real-time three-dimensional echocardiography was performed on each animal at baseline and at 1 hour and 8 weeks after infarction. Multiple valvular geometric measurements were calculated at each time point.
Results: Immediate and long-term changes in mitral valvular geometry were observed. Annular height–to–commissural width ratio decreased from 20.0% ± 1.6% to 11.2% ± 0.9% 1 hour after infarction (p < 0.001) and to 9.4% ± 0.4% 8 weeks after infarction (p < 0.001), whereas mitral annular area increased from 8.1 ± 0.3 cm2 to 9.2 ± 0.4 cm2 (p < 0.05) and then to 10.5 ± 0.6 cm2 (p < 0.05). Maximum mitral valve tenting area increased from 49.7 ± 5.1 mm2 to 58.6 ± 4.2 mm2 (p < 0.05) and then to 106.4 ± 3.9 mm2 (p < 0.001), whereas mitral valve tenting volume increased from 679.0 ± 75.5 mm3 to 828.6 ± 102.4 mm3 (p = 0.050) and then to 1530.5 ± 97.8 mm3 (p < 0.001). The mitral valve tenting index increased from 0.83 ± 0.08 mm to 0.88 ± 0.08 mm (p > 0.05) and then to 1.46 ± 0.08 mm (p < 0.001).
Conclusions: We have described a technique that uses real-time three-dimensional echocardiography to perform a comprehensive assessment of leaflet tethering on the entire mitral valve. Our methodology is not influenced by viewing plane selection, regional tenting asymmetry, or annular dilatation and therefore represents a potentially useful, clinically relevant, and consistent measure of subvalvular remodeling.
| Introduction |
|---|
|
|
|---|
During the past decade, laboratory and clinical studies using several different imaging modalities, including sonomicrometry [7], marker-tagged fluoroscopy [8], echocardiography [9], and magnetic resonance imaging (MRI) [10], have helped to elucidate the mechanisms that underlie IMR. These studies have demonstrated definitively that the two key anatomic elements contributing to the development of IMR are annular remodeling and disruption of the dynamic anatomic relationship between the subvalvular apparatus and the mitral annulus, which is manifest as systolic tethering of the leaflets into the left ventricular (LV) cavity. The extent to which each of these two elements contributes to valvular incompetence varies substantially among individual patients [11].
Although several indexes of annular remodeling, including mitral annular area, intercommissural width, and septolateral annular diameter, are readily quantified by standard two-dimensional (2D) echocardiographic techniques, methods for quantifying leaflet tethering have not been standardized. Descriptive geometric indicators derived from single-plane measurements made from 2D echocardiographic studies, such as tenting height and area, have been proposed [11]. Unfortunately, these measurements are inherently inaccurate because of the high variability of scanning planes and regional variation in both mitral annular shape and leaflet tenting [12–14] that develop in conjunction with postinfarction LV remodeling.
Recent advances in three-dimensional (3D) echocardiography have led some authors to propose multiplane leaflet tenting area [12] and leaflet tenting volume as indicators of subvalvular remodeling [14, 15]. Both tenting area and volume are, however, strongly influenced by annular size; therefore, neither index constitutes a reliable independent indicator of subvalvular remodeling. In this study, we report a methodology that allows for the characterization of leaflet tenting across the entire mitral valve surface as well as an index of 3D leaflet tenting that is independent of annular geometry.
| Material and Methods |
|---|
|
|
|---|
Ten sheep were pretreated with buprenorphine (2 µg/kg) and then induced with intravenous (IV) sodium thiopental (10 to 15 mg/kg), intubated, and anesthetized with isoflurane (1.5% to 2.0%) and oxygen (Narkomed, North American Drager, Telford, PA). All animals received glycopyrrolate (0.02 mg/kg IV) and cefazolin (1.0 g IV). The electrocardiogram, arterial blood pressure, and pulmonary artery pressure were monitored throughout the procedure.
A left thoracotomy was performed and baseline data were acquired. A posterobasal MI was then induced by the ligation of all branches of the circumflex coronary artery between the posterior ventricular vein and the posterior coronary artery. Data acquisition was repeated 1 hour after infarction. Eight weeks after infarction, the sheep were again anesthetized and monitored as previously described. Final data acquisition was performed after a limited upper midline laparotomy. All animals were euthanized after data acquisition.
Data Collection Protocol
Epicardial real-time 3D echocardiography was performed through a left thoracotomy immediately before and 1 hour after MI. Transdiaphragmatic echocardiography was performed through an upper midline laparotomy at the 8-week follow-up. In all cases, electrocardiogram-gated, full-volume data sets were acquired by a single operator using a Sonos 7500 (Philips Medical Systems, Andover, MA) platform equipped with a 2–4 MHz X4 handheld transducer. The degree of MR was determined quantitatively by assessing the area of the regurgitant jet as a percentage of left atrial area in the apical four-chamber view. The following grading was used: grade 1, less than 20%; grade 2, 20% to 40%; grade 3, 40% to 60%; and grade 4, more than 60% [16].
Image Analysis
Full-volume data sets were exported to a QLab 3D Advanced Quantification Software (Philips Medical Systems) workstation for calculation of LV cavity volumes. For each data set, endocardial contours were manually traced in both end diastolic and end systolic frames, which were selected by visual inspection. The endocardial contours of the remaining frames were traced in sequence by means of automated contour detection. Left ventricular end-diastolic (LVEDV) and end-systolic volumes (LVESV) were derived from the resultant global time-volume curve for each data set. The corresponding stroke volume and ejection fraction were both subsequently calculated.
Each full-volume data set was exported to a separate TomTec Cardio-View workstation (TomTec Imaging Systems, Munich, Germany) for geometric analysis. All calculations were performed at end systole. Image analysis was performed in Cardio-View by visual inspection. Cardio-View allows the interactive manipulation of 3D ultrasound data sets, including rotation, translation, surface rendering, and measurement. A rotational template consisting of 18 long-axis cross-sectional planes separated by 10° increments was created. This template was then centered at the geometric center of the mitral valve orifice and aligned with the axes of the mitral orifice. The two annular points intersecting each of the 18 long axis rotational planes were then marked interactively in Cardio-View.
Each data set was then exported to a Matlab (The Mathworks, Natick, MA) environment. The center of gravity of the resultant 36-point data set was translated to the origin. The least squares plane of the 3D data set was then calculated by means of orthogonal distance regression and the annular model rotated such that this mitral valve orifice plane was aligned with the x-y plane. Under these geometric conditions, the z coordinate (zn) of each annular point was equal to its distance to the plane of the mitral valve.
The Cartesian coordinates for each point in a given annular model were converted to cylindrical coordinates (r,
,z), and the data set rotated around the mitral valve axis (
direction) so that the data point containing zmax was located at
= 0 rad. Values of
for each point in a given data set were then recalculated in this fixed frame of reference.
Annular height (AH) for a given data set was then defined as zmax – zmin. Septolateral diameter for a given data set was defined as the distance, in 3D space, separating the two data points containing
= 0 rad and
=
rad. The intercommissural width (CW) for a given data set was defined as the distance, in 3D space, separating the data point with the lowest corresponding z value between 0 rad and
rad from the data point with the lowest corresponding z value between
rad and 2
rad. Annular height to commissural width ratio (AHCWR) was subsequently defined as AH/CW x 100%. Mitral annular area (MAA) was defined as the area enclosed by the 2D projection of a given annular data set onto its corresponding least squares plane. The spatial relationships of these geometric indexes are illustrated in Figure 1.
|
c, along the entire length of the intercommissural axis (Fig 2,C). To allow for interanimal comparison, intercommissural position was expressed as a percentage of the distance traveled between the anterior commissure and the posterior commissure, where 0% indicated the position of the anterior commissure and 100% indicated the position of the posterior commissure. The value of MVTa at the midpoint of the intercommissural axis (MVTamid) was recorded for each data set. The magnitude and the intercommissural position of the maximum segmental MVTa (MVTamax and PosMVTamax respectively) were recorded for each animal.
|
cn, was calculated. The mitral valve tenting index (MVTI), which was defined as MVTv/MAA, was calculated for each data set.
Statistical Analyses
For comparison of echocardiographic measurements between time points, two-tailed, paired t tests were used. For all comparisons, p
0.05 was considered significant. All statistical analysis was performed using SPSS software (SPSS Inc, Chicago, IL).
| Results |
|---|
|
|
|---|
|
|
Significant annular dilatation developed in all animals in conjunction with ventricular remodeling. Commissural width increased from 29.0 ± 1.3 mm at baseline to 35.7 ± 1.7 mm at 1 hour post-MI (p < 0.05) and 40.7 ± 1.5 mm at 8 weeks post-MI (p < 0.001). Septolateral diameter increased from 28.2 ± 0.8 mm to 30.2 ± 2.0 mm (p > 0.05) and then to 33.4 ± 2.3 mm (p < 0.05) over the same intervals. Mitral annular area increased from 8.1 ± 0.3 cm2 at baseline to 9.2 ± 0.4 cm2 at 1 hour post-MI (p < 0.05) and to 10.5 ± 0.6 cm2 at 8 weeks post-MI (p < 0.05).
Acute annular flattening was confirmed by a decrease in AH from 5.8 ± 0.5 mm at baseline to 4.0 ± 0.4 mm at 1 hour post-MI (p < 0.001). AHCWR decreased from 20.0% ± 1.6% to 11.2% ± 0.9% (p < 0.001) during the same interval. AH decreased to 3.9 ± 0.2 mm at 8 weeks post-MI (p < 0.05), and AHCWR decreased to 9.4% ± 0.4% (p < 0.001).
Mitral leaflet tethering was reflected by increases in multiple geometric measures during the observation period. MVTamax increased from 49.7 ± 5.1 mm2 at baseline to 58.6 ± 4.2 mm2 at 1 hour post-MI (p < 0.05) and to 106.4 ± 3.9 mm2 at 8 weeks post-MI (p < 0.01). Of interest was that the position of maximum tenting area shifted from near the midpoint of the intercommissural axis towards the anterior commissure as remodeling progressed: PosMVTamax decreased from 52.2% ± 0.9% at baseline to 41.6% ± 2.6% at 1 hour post-MI (p < 0.05) and to 34.3% ± 2.5% at 8 weeks post-MI (p < 0.001). Both increased leaflet tenting and progressive tenting asymmetry can be appreciated in Figure 3: the mean tenting area ± SEM for the cohort has been plotted as a function of intercommissural position at each of the three observation time points.
|
|
| Comment |
|---|
|
|
|---|
During the past decade, an increasingly aggressive surgical approach to IMR has developed. Mitral ring annuloplasty, with or without coronary revascularization, has become the preferred approach [20–22]. However, recent data suggest that the relief from recurrent MR afforded by mitral repair for IMR may be far less durable than initially surmised [23]. This lack of durability may be responsible for the difficulty in demonstrating a survival advantage for mitral valve surgical repair compared with medical management or revascularization alone in patients with IMR [24].
It is now recognized that the development of valvular incompetence after MI involves multiple geometric perturbations, including annular flattening and dilatation [25] as well as progressive subvalvular remodeling resulting in leaflet tethering [26]. Although mitral annuloplasty has the potential to restore normal annular size, it is less likely to address significant leaflet tethering. Failure to identify patients with extensive leaflet tethering preoperatively is likely a contributing factor to the disappointing operative outcomes that have been recently reported for annuloplasty in IMR patients [23].
Calafiore and colleagues [11] have proposed a relationship between increased 2D mitral leaflet coaptation depth and suboptimal operative and clinical outcomes after mitral valve repair for IMR [11]. These authors have proposed coaptation depth as a surrogate means of determining the extent of subvalvular remodeling present in a given patient and as a basis for selecting between valvular repair and replacement [27]. Recent studies, however, have shown that leaflet tenting is asymmetric and that the magnitude and pattern of asymmetry vary with infarct anatomy [12, 14]. These data suggest that isolated single-plane indexes of leaflet tenting as measured in either a parasternal long-axis or apical four-chamber view are potentially inaccurate, inconsistent, and inadequate to describe the pathologic anatomy and mechanics.
To address this potential shortcoming, Watanabe and colleagues [14, 15] have developed a 3D method for quantifying leaflet tenting volume and quantifying and localizing the site of peak tenting length (height). Although their findings are compelling in many regards, we believe their approach does not reliably reflect the extent of subvalvular remodeling for the following reasons:
In this study, we describe a technique that uses real-time 3D echocardiography to characterize (1) regional leaflet tenting across the entire surface of the mitral valve, (2) leaflet tenting asymmetry, and (3) leaflet tenting volume. Furthermore, we simultaneously quantify each of the three geometric perturbations thought to contribute to the development of IMR: annular dilatation, annular flattening, and leaflet tenting. Finally, we propose the mitral valve tenting index as a measure of subvalvular remodeling that is fully independent of annular remodeling.
Our methodology is not influenced by viewing plane selection or regional tenting asymmetry. The ratio that we describe, the mitral valve tenting index, allows for a quantitative description of leaflet tenting that is independent of mitral annular dilation. We believe that this index will serve as a means to reliably quantify the extent of leaflet tethering and act as a surrogate measure for the degree of subvalvular remodeling.
The ovine model used in this experiment is very reproducible and associated with consistent degrees of both annular and subvalvular remodeling. The high degree of reproducibility inherent in this model did not allow a correlative relationship between mitral valve tenting index and the degree of IMR to be established. In patients, the relative contributions of annular and subvalvular remodeling to the pathogenesis of IMR will be more variable. The ability to reliably quantify the relative contributions of annular and subvalvular remodeling in individual patients will allow better preoperative identification of those patients unlikely to achieve lasting relief of MR from annuloplasty alone.
However, before the mitral valve tenting index can be applied for clinical decision-making, we must validate the concept in serial studies of patients having surgical intervention for IMR who have variable degrees of MR that result from different infarct patterns (ie, anterior versus posterior walls). In the future, it may be possible that the ideal surgical approach for IMR will be individually tailored for each patient after an assessment of the 3D geometry and function of each component of the mitral valve complex [28].
Of interest is that our data demonstrate that the site of maximum leaflet tenting area shifts from the middle point of the intercommissural axis toward the anterior commissure during remodeling. This finding confirms that single-plane based indices of leaflet tenting (as determined by means of standard 2D echocardiographic techniques, such as MVTa and MVTh) are potentially unreliable.
The progressive anterior displacement of the location of peak leaflet tenting suggests that remodeling of the anterior papillary muscle is an important contributor to leaflet tenting in this model. This is counterintuitive given the posterobasal location of the infarction; however, previous studies performed in our laboratory help to explain this interesting finding. IMR in this model, as in humans, develops concomitantly with global LV dilation, which in turn influences the relationship between both papillary muscle groups and the annulus. Using sonomicrometry array localization, we have previously documented substantial increases in the distance between the tip of the anterior papillary muscle and the anterior as well as the posterior commissures [7], changes that are consistent with the pattern of leaflet tenting demonstrated in this experiment. These findings imply that the contribution of the anterior papillary muscle to leaflet tenting in IMR is significant and may have implications for evolving surgical approaches that focus solely on restoring normal posterior papillary-annular geometry [29, 30].
| Acknowledgments |
|---|
|
|
|---|
| References |
|---|
|
|
|---|
This article has been cited by other articles:
![]() |
T. A. Armen, R. Vandse, J. A. Crestanello, S. V. Raman, K. M. Bickle, and N. S. Nathan Mechanisms of valve competency after mitral valve annuloplasty for ischaemic mitral regurgitation using the Geoform ring: insights from three-dimensional echocardiography Eur J Echocardiogr, January 1, 2009; 10(1): 74 - 81. [Abstract] [Full Text] [PDF] |
||||
![]() |
S. Gelsomino, R. Lorusso, S. Caciolli, I. Capecchi, C. Rostagno, M. Chioccioli, G. De Cicco, G. Bille, P. Stefano, and G. F. Gensini Insights on left ventricular and valvular mechanisms of recurrent ischemic mitral regurgitation after restrictive annuloplasty and coronary artery bypass grafting. J. Thorac. Cardiovasc. Surg., August 1, 2008; 136(2): 507 - 518. [Abstract] [Full Text] [PDF] |
||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| ANN THORAC SURG | ASIAN CARDIOVASC THORAC ANN | EUR J CARDIOTHORAC SURG |
| J THORAC CARDIOVASC SURG | ICVTS | ALL CTSNet JOURNALS |