- Email: [email protected]
Accurate Localization of Mitral Regurgitant Defects Using Multiplane Transesophageal Echocardiography
Accurate Localization of Mitral Regurgitant Defects Using Multiplane Transesophageal Echocardiography Gary P. Foster, MD, Eric M. Isselbacher, MD, Geoffrey A. Rose, MD, David F. Torchiana, MD, Cary W. Akins, MD, and Michael H. Picard, MD Cardiac and Cardiac Surgical Units, Massachusetts General Hospital and Harvard Medical School, Boston, Massachusetts
Background. Appropriate patient selection for surgical repair of the mitral valve depends on the specific location and mechanism of regurgitation, which, in turn, has necessitated a more detailed method to accurately describe mitral pathology. This study tests a strategy of using multiplane transesophageal echocardiography to systematically localize mitral regurgitant defects and compares these results with the surgical findings. Methods. Fifty patients with mitral regurgitation underwent intraoperative transesophageal echocardiography for the evaluation of mitral pathology and potential repair. Mitral regurgitant defects were localized using a systematic strategy and a simple nomenclature that divides each mitral valve into six sections (three sections per leaflet) and each prosthetic sewing ring into six sections (60 radial degrees 5 one section). Results. Thirty-nine patients with native mitral valves were studied, for a total of 234 sections evaluated. Eightyseven of these sections contained regurgitant defects by
transesophageal echocardiography (mean number of regurgitant defects per valve, 2.2; range, 1 through 6). There was agreement between the transesophageal echocardiographic and surgical localizations in 96% (224/234; p < 0.0001) of the sections. Eleven patients with prosthetic mitral valves were studied, for a total of 66 sections evaluated. Twenty-three of these sections contained paravalvular leaks by transesophageal echocardiography (mean number of leaks per prosthesis, 2.1; range, 1 through 6). There was agreement between the transesophageal echocardiographic and surgical localizations in 88% (58/66; p < 0.001) of the sections. Conclusions. This transesophageal echocardiographic strategy provides a systematic method to accurately localize mitral regurgitant lesions and has the potential to improve the preoperative assessment of patients with significant mitral regurgitation. (Ann Thorac Surg 1998;65:1025–31) © 1998 by The Society of Thoracic Surgeons
S
often differ substantially. The surgeon views the mitral valve from the left atrium with the heart rotated to allow optimal exposure, whereas transesophageal images are obtained from multiple planes and perspectives. The surgeon also views the mitral valve in a nonphysiologic state because the heart has been decompressed during cardiopulmonary bypass, whereas transesophageal echocardiography images the mitral valve in its normal dynamic and physiologic state prior to cardiopulmonary bypass. Third, the currently used nomenclature for the anatomic description of the mitral valve is neither standardized nor universally accepted [11]. The purpose of this study was to develop a nomenclature and a strategy to systematically localize mitral regurgitant defects in both native and prosthetic valves using multiplane transesophageal echocardiography and to test the accuracy of this approach by comparison with the available reference standard of the surgical findings.
urgical repair of the mitral valve has become the treatment of choice in select patients with clinically significant mitral regurgitation. Because appropriate patient selection for valve repair depends on the specific location and mechanism of regurgitation, a more detailed preoperative description of mitral pathology has become increasingly necessary [1–3]. The imaging capabilities of transesophageal echocardiography make it ideally suited to assess mitral anatomy and pathology [4 –10]. This approach should allow regurgitant defects to be reproducibly located so that they can be followed throughout the course of the disease. Modern multiplane transducers image through 180 degrees of rotation while the transducer tip remains in a fixed position, thus allowing precise image alignment with the potential for accurate localization of pathologic defects. Despite these capabilities, routine transesophageal echocardiographic evaluation of the mitral valve remains challenging for several reasons. First, a systematic examination using standardized views is needed. Second, surgical and transesophageal echocardiographic descriptions of mitral anatomy and associated abnormalities
Accepted for publication Oct 31, 1997. Address reprint requests to Dr Foster, Cardiology Consultants, PC, 520 Medical Center Dr, Suite 100, Medford, OR 97504.
© 1998 by The Society of Thoracic Surgeons Published by Elsevier Science Inc
Material and Methods Patient Selection All patients with significant mitral regurgitation undergoing intraoperative transesophageal echocardiography in conjunction with valve repair or replacement from 0003-4975/98/$19.00 PII S0003-4975(98)00084-8
1026
FOSTER ET AL MITRAL DEFECTS BY TEE
1993 through 1996 at Massachusetts General Hospital were evaluated for inclusion in the study. The patients had either native or prosthetic mitral valve disease. Patients were excluded from the study if any of the following variables was present: they had major concomitant mitral stenosis, the transesophageal examination was performed with a monoplane or biplane transducer, the transesophageal echocardiographic study was incomplete, or the surgical reports were of insufficient detail to localize valvular pathology.
Mapping of Native Valve Pathology To assist in the localization of native mitral regurgitant defects, a reference diagram of the mitral valve was developed, similar to one proposed by Kumar and associates [11]. The reference view was chosen to reflect the anatomic perspective of looking from the left ventricular apex toward the base of the heart (Fig 1A). On this diagram, the mitral leaflets were each divided into three sections labeled in relation to their proximity to adjacent anatomic landmarks (aortic root, left atrial appendage, and papillary muscles): A1 through A3 for the anterior leaflet and P1 through P3 for the posterior leaflet. A1 and P1 represent the anterolateral sections; A2 and P2, the middle sections; and A3 and P3, the posteromedial sections. Lines indicating the transesophageal echocardiographic imaging planes were overlaid onto this diagram to indicate the specific mitral anatomy displayed in each plane (Fig 1B). The degrees of multiplane transducer rotation are shown next to each imaging plane as are the typical two-dimensional echocardiographic images produced at each rotational position. To display the same mitral anatomy as it would be seen from the typical surgical perspective, this diagram was horizontally flipped around a vertical axis (mirror image) and then rotated 45 degrees counterclockwise (Fig 1C).
Mapping of Prosthetic Valve Pathology Using an approach similar to that just described, a second diagram was devised to describe the anatomic relationships of prosthetic mitral valves (Fig 2A). Again, the diagram could be horizontally flipped and then rotated to provide the corresponding surgical view (Fig 2B). In the surgical view, the circumference of the sewing ring was then overlaid with a clock face and oriented in such a way that the aorta was located at the 12 o’clock position.
Imaging Strategy for Native Valves A Hewlett-Packard multiplane transesophageal echocardiographic probe was used in conjunction with either a Hewlett-Packard Sonos 1000 or 1500 Ultrasound Imaging System (Andover, MA). Images were recorded on 0.5inch VHS videotapes for blinded and independent review. To perform a comprehensive examination of the mitral valve, it is essential to understand how transesophageal probe maneuvers change the imaging plane with respect to the mitral valve. Our typical examination begins with
Ann Thorac Surg 1998;65:1025–31
an assessment of the valve in three horizontal planes (Fig 3A). In the standard midesophageal four-chamber view (typically at a transducer angle of 0 degrees), the middle portions of both leaflets (A2 and P2) are visualized. From this position, flexion, withdrawal, or both of the transducer tip allow visualization of the aortic root and the anterolateral portions of the mitral leaflets (A1 and P1). Similarly, retroflexion, advancement, or both of the transducer tip allow visualization of the posteromedial portions of the leaflets (A3 and P3). To confirm the locations of mitral leaflet pathology identified in these views, the mitral leaflets are examined a second time with the imaging plane parallel to the major axis of the mitral orifice, which is typically obtained from the midesophageal position by adjusting the transducer plane to an angle of between 45 and 90 degrees. When centered through the major axis of the valve orifice, this view allows evaluation of P1, A2, and P3 (Fig 3B). By manual rotation of the probe shaft in a clockwise direction, the entire anterior leaflet is visualized (A1, A2, and A3). Next, by counterclockwise probe rotation, the entire posterior leaflet is visualized (P1, P2, and P3). Other imaging planes and transducer positions were used when available, such as the transgastric shortaxis view, to assist in confirming the location of the regurgitant defect. Using the strategy outline here, each leaflet section was examined for the presence of prolapse or ruptured chordae tendineae and the site of mitral regurgitation. If other regurgitant defects were present (eg, fenestrations), their location was identified in this manner as well. An example of focal P1 prolapse is shown in Figure 4.
Imaging Strategy for Prosthetic Valves Similar principles apply to the mapping of prosthetic valve regurgitant lesions, ie, using a series of sequential views to visualize all segments of the mitral valve and sewing ring. The prosthetic sewing ring is centered in the standard midesophageal four-chamber view. The sewing ring is then imaged in full by rotation of the imaging plane through successive 10-degree increments while the probe and transducer tip are adjusted as necessary to keep the plane centered through the sewing ring. The attachment of the sewing ring to the mitral annulus is inspected using both two-dimensional and color Doppler imaging through the full 180 degrees of rotation, thus allowing the entire circumference of the sewing ring to be examined. To simplify the comparison with surgical findings, the clockface was divided into six circumferential sections of 60 degrees each (ie, 2 hours). An example of a single paravalvular leak located at the 7 o’clock position is shown in Figure 5.
Reference Standard and Analysis After completion of the mitral valve or replacement procedure, a detailed description of the location and characteristics of the mitral pathology was provided by the surgeon as recorded in the operative report. This operative report then served as the reference standard
Ann Thorac Surg 1998;65:1025–31
FOSTER ET AL MITRAL DEFECTS BY TEE
1027
against which the transesophageal echocardiographic imaging strategies were tested. The videotape for each study was reviewed off line and independently by one of three trained echocardiographers blinded to the surgical results. The locations of all regurgitant defects were then plotted on the reference diagrams already described (see Fig 1A, 2A). The regurgitant defect locations determined by transesophageal echocardiography and at operation were then compared.
Statistical Methods The transesophageal echocardiographic and surgical examinations of all native and prosthetic mitral valves were compared on a section by section basis. The total number of sections in which there was agreement between the transesophageal echocardiographic and surgical findings was compared with the total number of mitral valve sections evaluated in the study population using Fisher’s test, and a p value of less than 0.05 was considered significant.
Fig 1. (A) Reference view displaying the mitral valve and its anatomic relationship to the aortic root (Ao) and the left atrial appendage (LAA) as seen from the left ventricular apex. (B) Reference view demonstrating the relationship of the transesophageal echocardiographic imaging planes to the mitral valve with the probe positioned in the standard midesophageal position. (C) Surgical view of the mitral valve as seen from left atrium with the heart rotated. See text for details. (A1, A2, A3 5 anterior leaflet sections; P1, P2, P3 5 posterior leaflet sections.)
Fig 2. (A) Reference view displaying the prosthetic mitral valve and its anatomic relationship to the aortic root (Ao) and left atrial appendage (LAA) as seen from the left ventricular apex. The hours of a clock face, corresponding to those shown in the surgical perspective, have been overlaid. (B) Surgical view of prosthetic mitral valve and its relationship to the aortic root.
1028
FOSTER ET AL MITRAL DEFECTS BY TEE
Ann Thorac Surg 1998;65:1025–31
Fig 3. (A) Effect of flexion or withdrawal and retroflexion or advancement of the transesophageal probe tip on the imaging plane in relation to the mitral valve at a transducer rotational angle of 0 degrees. (B) Effect of clockwise and counterclockwise probe rotation on the imaging plane in relation to the mitral valve with the transducer rotational angle adjusted to the major axis of the mitral orifice (typically 45 to 90 degrees). (Ao 5 aorta; A1, A2, A3 5 anterior leaflet sections; LAA 5 left atrial appendage; P1, P2, P3 5 posterior leaflet sections.)
Results Native Mitral Regurgitant Defects A total of 39 patients with native mitral valves were studied to localize mitral pathology and regurgitant de-
fects. As each mitral valve had six sections, a total of 234 mitral sections were evaluated. Of these 234 sections, 82 were found by transesophageal echocardiography to have regurgitant defects. In all 39 patients, mitral regurgitation was the result of prolapsing or flail leaflet seg-
Ann Thorac Surg 1998;65:1025–31
FOSTER ET AL MITRAL DEFECTS BY TEE
1029
Fig 4. Example of a focal P1 prolapse identified at a transducer rotational angle of 0 degrees with the imaging plane through the aortic root (AR) and the anterolateral portion of the mitral leaflets. (LA 5 left atrium; LV 5 left ventricle; RV 5 right ventricle.)
ments. A mean of 2.2 regurgitant defects per valve (range, 1 through 6) were identified. The majority of these regurgitant defects were found in the posterior leaflet (Table 1). The anterior leaflet alone was involved in 6 patients, the posterior leaflet alone in 23, and both leaflets in 10 patients. For both the anterior and posterior leaflets, the middle sections (A2 and P2) were most commonly involved. There was agreement between the transesophageal echocardiographic and surgical findings in 224 (96%) (p , 0.0001) of the 234 sections examined. In each of the 10 where there was disagreement, the regurgitant defect was localized to an adjacent valve section.
Prosthetic Mitral Regurgitant Defects A total of 11 patients with prosthetic mitral valves were examined for the localization of paravalvular leaks. Six sections of 60 radial degrees were examined on each prosthesis, for a total of 66 sections available for analysis. Of these 66 sections, 21 were found by transesophageal echocardiography to have paravalvular leaks. A mean of 1.9 leaks per prosthesis (range, 1 through 6) were identi-
fied. Paravalvular leak location was fairly evenly distributed around the circumference of the sewing ring in our study population (Table 2). There was agreement between the findings of transesophageal echocardiography and operation in 58 (88%) (p , 0.001) of the 66 sections examined. In the eight where there was disagreement, the cause was localization of a leak to adjacent sections of the prosthetic sewing ring. In the remaining two, one leak appeared by color Doppler to be extensive but was found at operation to originate from a discrete point, and the other involved a secondary paravalvular leak. It was noted on the preoperative transesophageal echocardiographic study and, although not identified by the surgeon intraoperatively, persisted on the postoperative transesophageal echocardiographic study.
Comment This study demonstrates the use of a strategy to localize mitral regurgitant defects using multiplane transesophageal echocardiography. This approach is shown to be
Fig 5. Example of a discrete paravalvular leak (PVL) at a transducer rotational angle of 0 degrees corresponding to a location of 7 o’clock on the surgical clock face. (BPMV 5 bioprosthetic mitral valve; LA 5 left atrium; LV 5 left ventricle.)
1030
FOSTER ET AL MITRAL DEFECTS BY TEE
Ann Thorac Surg 1998;65:1025–31
Table 1. Native Mitral Regurgitant Defect Location as Determined by Transesophageal Echocardiographya No. of Defects (n 5 82)
Defect Location A1 A2 A3 P1 P2 P3 a
6 (7) 13 (16) 6 (7) 18 (22) 26 (32) 13 (16)
Numbers in parentheses are percentages.
A1, A2, A3 5 anterior leaflet sections; sections.
P1, P2, P3 5 posterior leaflet
accurate for both native and prosthetic mitral regurgitant defects when compared with the surgical findings. An improved understanding of the underlying pathophysiology of mitral regurgitation has led to advances in surgical therapies [12, 13]. Specifically, in selected patients, surgical repair has become the treatment of choice [14]. Indeed, valve repair is a particularly attractive option, as it avoids the morbidity associated with both valvular prostheses and long-term anticoagulation [14]. The mitral pathology most amenable to surgical repair involves disease localized to the posterior leaflet or to a focal portion of the anterior leaflet. Conversely, repair is less often successful when there is extensive disease of the anterior leaflet or when there is incomplete closure of the mitral valve, and in such cases, valve replacement is often required [1, 15, 16]. Advances in echocardiographic technology have improved the detail in which the mitral leaflets can be examined in real time. Although monoplane and biplane transesophageal echocardiographic transducers can visualize select portions of the mitral leaflets, limited imaging planes make a thorough assessment of the entire valve technically demanding. Accordingly, considerable expertise is necessary to obtain adequate information regarding the exact site of mitral leaflet pathology when such transducers are used. With the newer multiplane imaging probes, the entire mitral valve can be visualized with relatively little instrument manipulation, allowing adequate imaging in almost any given anatomic or physiologic variation. Advances in both surgical techniques and imaging technology have necessitated the development of a sysTable 2. Prosthetic Mitral Regurgitant Defect Location as Determined by Transesophageal Echocardiographya Leak Location 12:00 –2:00 2:00 – 4:00 4:00 – 6:00 6:00 – 8:00 8:00 –10:00 10:00 –12:00 a
Numbers in parentheses are percentages.
No. of Leaks (n 5 21) 3 (14) 4 (19) 4 (19) 4 (19) 2 (10) 4 (19)
tematic transesophageal echocardiographic approach for the evaluation of mitral regurgitation. To be clinically useful, valvular abnormalities must be accurately identified and their location defined relative to fixed anatomic landmarks. Several prior studies [4, 6, 8 –10, 17–22] have demonstrated the ability of transesophageal echocardiography to evaluate the mitral valve and identify its pathology. Fehske and colleagues [9] described the mitral pathologic findings in a group of patients in whom a complete examination of the leaflets was facilitated by the use of multiplane transesophageal echocardiography. Similarly, Stewart and co-workers [10] presented the advantages of multiplane transesophageal echocardiography in the complete examination of the mitral leaflets. Although these studies describe the mitral pathologic findings, they provide little information as to how one should systematically evaluate the valve. A natural instinct for the echocardiographer is to concentrate on conspicuous pathologic findings; however, the presence of leaflet sections with minor disease and the absence of pathologic findings are equally important, thus necessitating a complete and systematic examination. The current study proposes such a systematic strategy using multiplane transesophageal echocardiography and validates its efficacy against a reference standard. The distribution of defects of native mitral leaflets observed in this study was a function of the patients referred for mitral valve procedures. In our series, 70% of the lesions were localized to the posterior leaflet, a finding similar to that in a series of mitral valve repair for mitral regurgitation caused by degenerative disease [16]. In the current study, there were a small number of disagreements between the transesophageal echocardiographic and surgical findings in the native valve group. The majority of these disagreements were a consequence of localization of pathology to adjacent mitral valve sections and for the most part occurred when regurgitant lesions straddled the schematically defined sections. Similarly, in the prosthetic valve group, six of the eight disagreements were due to localization of paravalvular leaks to adjacent sections, ie, within 30 degrees between the two methods. Most of these disagreements also resulted from regurgitant jets that straddled two or more of the schematically defined sections. Of the two remaining cases of disagreement, one was an eccentrically directed paravalvular regurgitant jet that appeared echocardiographically to be a wide leak, but at operation was found to be a localized defect without extension. The surgical finding in this instance showed a focal defect with no extension. The other case was a paravalvular leak that was identified and localized preoperatively by transesophageal echocardiography but that the surgeon was unable to find once cardiopulmonary bypass was established; this paravalvular leak was still present on the postoperative transesophageal examination. This last case exemplifies the challenge surgeons have in trying to assess mitral defects in a nonphysiologic state. When the left ventricle is decompressed, the surrounding anatomy becomes distorted, making paravalvular leaks difficult to detect or localize. For native mitral
Ann Thorac Surg 1998;65:1025–31
valves, the surgeon is generally limited to assessment of the left atrial surface of the leaflets while the heart is decompressed, and this again leads to distortion of the mitral anatomy. Although insufflation of the left ventricle is a helpful technique in the evaluation of valvular and paravalvular competence, it remains a crude substitute for the true dynamic physiologic state. In contrast to the surgical setting, transesophageal echocardiography is able to evaluate the mitral valve in a normal dynamic physiologic situation without distortion of the surrounding cardiac structures. Thus, transesophageal echocardiographic findings more clearly reflect true valvular function. One last advantage of the strategy described in this study is that it provides a standardized nomenclature to localize mitral valve pathology and that this has the potential to improve communication between the surgeon and the echocardiographer. Moreover, this method should prove useful in defining the pathologic features of the mitral leaflets during preoperative transesophageal echocardiographic evaluation. This knowledge should allow decisions regarding mitral repair versus replacement to be made outside the operating room setting. The current study was limited in its ability to evaluate and compare prospectively the transesophageal echocardiographic and surgical findings without the introduction of bias. To limit this bias, the transesophageal echocardiographic studies were independently reviewed off line before comparison with the surgical findings as outlined in the detailed surgical report. Nonetheless, bias may still have been introduced during the performance of the intraoperative transesophageal echocardiogram, at which time cursory echocardiographic findings were relayed to the surgeon that could potentially have influenced the reporting of mitral pathology. A promising approach to localizing mitral pathology is through three-dimensional echocardiographic reconstruction of regurgitant defects and their associated color Doppler jets, but at present, this procedure is both time-consuming and incapable of giving “on-line” information [23]. The approach used in the current study provides a rapid and accurate method for the general echocardiographer to use until the time that on-line three-dimensional mitral valve reconstruction becomes available for surgical decision making. In conclusion, this transesophageal echocardiographic strategy provides a systematic method to accurately localize native and prosthetic mitral regurgitant lesions. By accurately localizing these defects, this strategy has the potential to improve the preoperative assessment, communication, and decisions about patients with significant mitral regurgitation.
References 1. Cosgrove DM, Stewart WJ. Mitral valvuloplasty. Curr Probl Cardiol 1989;14:359 – 415. 2. Dion R. Ischemic mitral regurgitation: when and how should it be corrected? J Heart Valve Dis 1993;2:536– 43. 3. Odell JA, Orszulak TA. Surgical repair and reconstruction of valvular lesions. Curr Opin Cardiol 1995;10:135– 43. 4. De Simone R, Lange R, Saggau W, Gams E, Tanzeem A, Hagl
FOSTER ET AL MITRAL DEFECTS BY TEE
5.
6.
7.
8.
9. 10. 11. 12.
13.
14. 15. 16. 17.
18.
19.
20.
21. 22. 23.
1031
S. Intraoperative transesophageal echocardiography for the evaluation of mitral, aortic and tricuspid valve repair. A tool to optimize surgical outcome. Eur J Cardiothorac Surg 1992; 6:665–73. Himelman RB, Kusumoto F, Oken K, et al. The flail mitral valve: echocardiographic findings by precordial and transesophageal imaging and Doppler color flow mapping. J Am Coll Cardiol 1991;17:272–9. Freeman WK, Schaff HV, Khandheria BK, et al. Intraoperative evaluation of mitral valve regurgitation and repair by transesophageal echocardiography: incidence and significance of systolic anterior motion. J Am Coll Cardiol 1992;20: 599 – 609. Homma S, Mullis S, Di Tullio MD, Schneller S, Apfelbaum M, Smith C. A transesophageal echocardiographic demonstration of ischemic mitral regurgitation. J Am Soc Echocardiogr 1993;6:341– 4. Pieper EP, Hellemans IM, Hamer HP, et al. Additional value of biplane transesophageal echocardiography in assessing the genesis of mitral regurgitation and the feasibility of valve repair. Am J Cardiol 1995;75:489–93. Fehske W, Grayburn PA, Omran H, et al. Morphology of the mitral valve as displayed by multiplane transesophageal echocardiography. J Am Soc Echocardiogr 1994;7:472–9. Stewart WJ, Griffin B, Thomas JD. Multiplane transesophageal echocardiographic evaluation of mitral valve disease. Am J Cardiac Imaging 1995;9:121– 8. Kumar N, Kumar M, Duran CM. A revised terminology for recording surgical findings of the mitral valve. J Heart Valve Dis 1995;4:70–5. Seward JB, Khanderia BK, Freeman WK, et al. Multiplane transesophageal echocardiography: image orientation, examination technique, anatomic correlations, and clinical applications. Mayo Clinic Proc 1993;68:523–51. Stewart WJ, Currie PJ, Salcedo EE, et al. Evaluation of mitral leaflet motion by echocardiography and jet direction by Doppler color flow mapping to determine the mechanism of mitral regurgitation. J Am Coll Cardiol 1992;20:1353– 61. Akins CW, Hilgenberg AD, Buckley MJ, et al. Mitral valve reconstruction versus replacement for degenerative or ischemic mitral regurgitation. Ann Thorac Surg 1994;58:668–76. Antunes MJ. Mitral valve repair into the 1990s. Eur J Cardiothorac Surg 1992;6:S13– 6. David TE, Armstrong S, Sun Z, Daniel L. Late results of mitral valve repair for mitral regurgitation due to degenerative disease. Ann Thorac Surg 1993;56:7–14. Hellemans IM, Pieper EG, Ravelli ACJ, et al. Comparison of transthoracic and transesophageal echocardiography with surgical findings in mitral regurgitation. Am J Cardiol 1996; 77:728–33. Flachskampf FA, Hoffmann R, Franke A, et al. Does multiplane transesophageal echocardiography improve the assessment of prosthetic valve regurgitation? J Am Soc Echocardiogr 1995;8:70– 8. Alam M, Serwin JB, Rosman HS, Polanco GA, Sun I, Silverman NA. Transesophageal echocardiographic features of normal and disfunctioning bioprosthetic valves. Am Heart J 1991;121:1149–55. Chaudhry FA, Herrera C, DeFrino PF, Mehlman DJ, Zabalgoitia M. Pathologic and angiographic correlations of transesophageal echocardiography in prosthetic heart valve dysfunction. Am Heart J 1991;122:1057– 64. Maurer G, Siegel RJ, Czer LSC. The use of color flow mapping for intraoperative assessment of valve repair. Circulation 1991;84(Suppl 1):250– 8. Meloni L, Aru G, Abbruzzese PA, Cardu G, Martelli V, Cherchi A. Localization of mitral periprosthetic leaks by transesophageal echocardiography. Am J Cardiol 1992;69:276–9. Salustri A, Becker AE, Van Herwerden L, Vletter WB, Ten Cate FJ, Roetlandt JRT. Three-dimensional echocardiography of normal and pathologic mitral valve: a comparison with two-dimensional transesophageal echocardiography. J Am Coll Cardiol 1996;27:1502–10.
Background. Appropriate patient selection for surgical repair of the mitral valve depends on the specific location and mechanism of regurgitation, which, in turn, has necessitated a more detailed method to accurately describe mitral pathology. This study tests a strategy of using multiplane transesophageal echocardiography to systematically localize mitral regurgitant defects and compares these results with the surgical findings. Methods. Fifty patients with mitral regurgitation underwent intraoperative transesophageal echocardiography for the evaluation of mitral pathology and potential repair. Mitral regurgitant defects were localized using a systematic strategy and a simple nomenclature that divides each mitral valve into six sections (three sections per leaflet) and each prosthetic sewing ring into six sections (60 radial degrees 5 one section). Results. Thirty-nine patients with native mitral valves were studied, for a total of 234 sections evaluated. Eightyseven of these sections contained regurgitant defects by
transesophageal echocardiography (mean number of regurgitant defects per valve, 2.2; range, 1 through 6). There was agreement between the transesophageal echocardiographic and surgical localizations in 96% (224/234; p < 0.0001) of the sections. Eleven patients with prosthetic mitral valves were studied, for a total of 66 sections evaluated. Twenty-three of these sections contained paravalvular leaks by transesophageal echocardiography (mean number of leaks per prosthesis, 2.1; range, 1 through 6). There was agreement between the transesophageal echocardiographic and surgical localizations in 88% (58/66; p < 0.001) of the sections. Conclusions. This transesophageal echocardiographic strategy provides a systematic method to accurately localize mitral regurgitant lesions and has the potential to improve the preoperative assessment of patients with significant mitral regurgitation. (Ann Thorac Surg 1998;65:1025–31) © 1998 by The Society of Thoracic Surgeons
S
often differ substantially. The surgeon views the mitral valve from the left atrium with the heart rotated to allow optimal exposure, whereas transesophageal images are obtained from multiple planes and perspectives. The surgeon also views the mitral valve in a nonphysiologic state because the heart has been decompressed during cardiopulmonary bypass, whereas transesophageal echocardiography images the mitral valve in its normal dynamic and physiologic state prior to cardiopulmonary bypass. Third, the currently used nomenclature for the anatomic description of the mitral valve is neither standardized nor universally accepted [11]. The purpose of this study was to develop a nomenclature and a strategy to systematically localize mitral regurgitant defects in both native and prosthetic valves using multiplane transesophageal echocardiography and to test the accuracy of this approach by comparison with the available reference standard of the surgical findings.
urgical repair of the mitral valve has become the treatment of choice in select patients with clinically significant mitral regurgitation. Because appropriate patient selection for valve repair depends on the specific location and mechanism of regurgitation, a more detailed preoperative description of mitral pathology has become increasingly necessary [1–3]. The imaging capabilities of transesophageal echocardiography make it ideally suited to assess mitral anatomy and pathology [4 –10]. This approach should allow regurgitant defects to be reproducibly located so that they can be followed throughout the course of the disease. Modern multiplane transducers image through 180 degrees of rotation while the transducer tip remains in a fixed position, thus allowing precise image alignment with the potential for accurate localization of pathologic defects. Despite these capabilities, routine transesophageal echocardiographic evaluation of the mitral valve remains challenging for several reasons. First, a systematic examination using standardized views is needed. Second, surgical and transesophageal echocardiographic descriptions of mitral anatomy and associated abnormalities
Accepted for publication Oct 31, 1997. Address reprint requests to Dr Foster, Cardiology Consultants, PC, 520 Medical Center Dr, Suite 100, Medford, OR 97504.
© 1998 by The Society of Thoracic Surgeons Published by Elsevier Science Inc
Material and Methods Patient Selection All patients with significant mitral regurgitation undergoing intraoperative transesophageal echocardiography in conjunction with valve repair or replacement from 0003-4975/98/$19.00 PII S0003-4975(98)00084-8
1026
FOSTER ET AL MITRAL DEFECTS BY TEE
1993 through 1996 at Massachusetts General Hospital were evaluated for inclusion in the study. The patients had either native or prosthetic mitral valve disease. Patients were excluded from the study if any of the following variables was present: they had major concomitant mitral stenosis, the transesophageal examination was performed with a monoplane or biplane transducer, the transesophageal echocardiographic study was incomplete, or the surgical reports were of insufficient detail to localize valvular pathology.
Mapping of Native Valve Pathology To assist in the localization of native mitral regurgitant defects, a reference diagram of the mitral valve was developed, similar to one proposed by Kumar and associates [11]. The reference view was chosen to reflect the anatomic perspective of looking from the left ventricular apex toward the base of the heart (Fig 1A). On this diagram, the mitral leaflets were each divided into three sections labeled in relation to their proximity to adjacent anatomic landmarks (aortic root, left atrial appendage, and papillary muscles): A1 through A3 for the anterior leaflet and P1 through P3 for the posterior leaflet. A1 and P1 represent the anterolateral sections; A2 and P2, the middle sections; and A3 and P3, the posteromedial sections. Lines indicating the transesophageal echocardiographic imaging planes were overlaid onto this diagram to indicate the specific mitral anatomy displayed in each plane (Fig 1B). The degrees of multiplane transducer rotation are shown next to each imaging plane as are the typical two-dimensional echocardiographic images produced at each rotational position. To display the same mitral anatomy as it would be seen from the typical surgical perspective, this diagram was horizontally flipped around a vertical axis (mirror image) and then rotated 45 degrees counterclockwise (Fig 1C).
Mapping of Prosthetic Valve Pathology Using an approach similar to that just described, a second diagram was devised to describe the anatomic relationships of prosthetic mitral valves (Fig 2A). Again, the diagram could be horizontally flipped and then rotated to provide the corresponding surgical view (Fig 2B). In the surgical view, the circumference of the sewing ring was then overlaid with a clock face and oriented in such a way that the aorta was located at the 12 o’clock position.
Imaging Strategy for Native Valves A Hewlett-Packard multiplane transesophageal echocardiographic probe was used in conjunction with either a Hewlett-Packard Sonos 1000 or 1500 Ultrasound Imaging System (Andover, MA). Images were recorded on 0.5inch VHS videotapes for blinded and independent review. To perform a comprehensive examination of the mitral valve, it is essential to understand how transesophageal probe maneuvers change the imaging plane with respect to the mitral valve. Our typical examination begins with
Ann Thorac Surg 1998;65:1025–31
an assessment of the valve in three horizontal planes (Fig 3A). In the standard midesophageal four-chamber view (typically at a transducer angle of 0 degrees), the middle portions of both leaflets (A2 and P2) are visualized. From this position, flexion, withdrawal, or both of the transducer tip allow visualization of the aortic root and the anterolateral portions of the mitral leaflets (A1 and P1). Similarly, retroflexion, advancement, or both of the transducer tip allow visualization of the posteromedial portions of the leaflets (A3 and P3). To confirm the locations of mitral leaflet pathology identified in these views, the mitral leaflets are examined a second time with the imaging plane parallel to the major axis of the mitral orifice, which is typically obtained from the midesophageal position by adjusting the transducer plane to an angle of between 45 and 90 degrees. When centered through the major axis of the valve orifice, this view allows evaluation of P1, A2, and P3 (Fig 3B). By manual rotation of the probe shaft in a clockwise direction, the entire anterior leaflet is visualized (A1, A2, and A3). Next, by counterclockwise probe rotation, the entire posterior leaflet is visualized (P1, P2, and P3). Other imaging planes and transducer positions were used when available, such as the transgastric shortaxis view, to assist in confirming the location of the regurgitant defect. Using the strategy outline here, each leaflet section was examined for the presence of prolapse or ruptured chordae tendineae and the site of mitral regurgitation. If other regurgitant defects were present (eg, fenestrations), their location was identified in this manner as well. An example of focal P1 prolapse is shown in Figure 4.
Imaging Strategy for Prosthetic Valves Similar principles apply to the mapping of prosthetic valve regurgitant lesions, ie, using a series of sequential views to visualize all segments of the mitral valve and sewing ring. The prosthetic sewing ring is centered in the standard midesophageal four-chamber view. The sewing ring is then imaged in full by rotation of the imaging plane through successive 10-degree increments while the probe and transducer tip are adjusted as necessary to keep the plane centered through the sewing ring. The attachment of the sewing ring to the mitral annulus is inspected using both two-dimensional and color Doppler imaging through the full 180 degrees of rotation, thus allowing the entire circumference of the sewing ring to be examined. To simplify the comparison with surgical findings, the clockface was divided into six circumferential sections of 60 degrees each (ie, 2 hours). An example of a single paravalvular leak located at the 7 o’clock position is shown in Figure 5.
Reference Standard and Analysis After completion of the mitral valve or replacement procedure, a detailed description of the location and characteristics of the mitral pathology was provided by the surgeon as recorded in the operative report. This operative report then served as the reference standard
Ann Thorac Surg 1998;65:1025–31
FOSTER ET AL MITRAL DEFECTS BY TEE
1027
against which the transesophageal echocardiographic imaging strategies were tested. The videotape for each study was reviewed off line and independently by one of three trained echocardiographers blinded to the surgical results. The locations of all regurgitant defects were then plotted on the reference diagrams already described (see Fig 1A, 2A). The regurgitant defect locations determined by transesophageal echocardiography and at operation were then compared.
Statistical Methods The transesophageal echocardiographic and surgical examinations of all native and prosthetic mitral valves were compared on a section by section basis. The total number of sections in which there was agreement between the transesophageal echocardiographic and surgical findings was compared with the total number of mitral valve sections evaluated in the study population using Fisher’s test, and a p value of less than 0.05 was considered significant.
Fig 1. (A) Reference view displaying the mitral valve and its anatomic relationship to the aortic root (Ao) and the left atrial appendage (LAA) as seen from the left ventricular apex. (B) Reference view demonstrating the relationship of the transesophageal echocardiographic imaging planes to the mitral valve with the probe positioned in the standard midesophageal position. (C) Surgical view of the mitral valve as seen from left atrium with the heart rotated. See text for details. (A1, A2, A3 5 anterior leaflet sections; P1, P2, P3 5 posterior leaflet sections.)
Fig 2. (A) Reference view displaying the prosthetic mitral valve and its anatomic relationship to the aortic root (Ao) and left atrial appendage (LAA) as seen from the left ventricular apex. The hours of a clock face, corresponding to those shown in the surgical perspective, have been overlaid. (B) Surgical view of prosthetic mitral valve and its relationship to the aortic root.
1028
FOSTER ET AL MITRAL DEFECTS BY TEE
Ann Thorac Surg 1998;65:1025–31
Fig 3. (A) Effect of flexion or withdrawal and retroflexion or advancement of the transesophageal probe tip on the imaging plane in relation to the mitral valve at a transducer rotational angle of 0 degrees. (B) Effect of clockwise and counterclockwise probe rotation on the imaging plane in relation to the mitral valve with the transducer rotational angle adjusted to the major axis of the mitral orifice (typically 45 to 90 degrees). (Ao 5 aorta; A1, A2, A3 5 anterior leaflet sections; LAA 5 left atrial appendage; P1, P2, P3 5 posterior leaflet sections.)
Results Native Mitral Regurgitant Defects A total of 39 patients with native mitral valves were studied to localize mitral pathology and regurgitant de-
fects. As each mitral valve had six sections, a total of 234 mitral sections were evaluated. Of these 234 sections, 82 were found by transesophageal echocardiography to have regurgitant defects. In all 39 patients, mitral regurgitation was the result of prolapsing or flail leaflet seg-
Ann Thorac Surg 1998;65:1025–31
FOSTER ET AL MITRAL DEFECTS BY TEE
1029
Fig 4. Example of a focal P1 prolapse identified at a transducer rotational angle of 0 degrees with the imaging plane through the aortic root (AR) and the anterolateral portion of the mitral leaflets. (LA 5 left atrium; LV 5 left ventricle; RV 5 right ventricle.)
ments. A mean of 2.2 regurgitant defects per valve (range, 1 through 6) were identified. The majority of these regurgitant defects were found in the posterior leaflet (Table 1). The anterior leaflet alone was involved in 6 patients, the posterior leaflet alone in 23, and both leaflets in 10 patients. For both the anterior and posterior leaflets, the middle sections (A2 and P2) were most commonly involved. There was agreement between the transesophageal echocardiographic and surgical findings in 224 (96%) (p , 0.0001) of the 234 sections examined. In each of the 10 where there was disagreement, the regurgitant defect was localized to an adjacent valve section.
Prosthetic Mitral Regurgitant Defects A total of 11 patients with prosthetic mitral valves were examined for the localization of paravalvular leaks. Six sections of 60 radial degrees were examined on each prosthesis, for a total of 66 sections available for analysis. Of these 66 sections, 21 were found by transesophageal echocardiography to have paravalvular leaks. A mean of 1.9 leaks per prosthesis (range, 1 through 6) were identi-
fied. Paravalvular leak location was fairly evenly distributed around the circumference of the sewing ring in our study population (Table 2). There was agreement between the findings of transesophageal echocardiography and operation in 58 (88%) (p , 0.001) of the 66 sections examined. In the eight where there was disagreement, the cause was localization of a leak to adjacent sections of the prosthetic sewing ring. In the remaining two, one leak appeared by color Doppler to be extensive but was found at operation to originate from a discrete point, and the other involved a secondary paravalvular leak. It was noted on the preoperative transesophageal echocardiographic study and, although not identified by the surgeon intraoperatively, persisted on the postoperative transesophageal echocardiographic study.
Comment This study demonstrates the use of a strategy to localize mitral regurgitant defects using multiplane transesophageal echocardiography. This approach is shown to be
Fig 5. Example of a discrete paravalvular leak (PVL) at a transducer rotational angle of 0 degrees corresponding to a location of 7 o’clock on the surgical clock face. (BPMV 5 bioprosthetic mitral valve; LA 5 left atrium; LV 5 left ventricle.)
1030
FOSTER ET AL MITRAL DEFECTS BY TEE
Ann Thorac Surg 1998;65:1025–31
Table 1. Native Mitral Regurgitant Defect Location as Determined by Transesophageal Echocardiographya No. of Defects (n 5 82)
Defect Location A1 A2 A3 P1 P2 P3 a
6 (7) 13 (16) 6 (7) 18 (22) 26 (32) 13 (16)
Numbers in parentheses are percentages.
A1, A2, A3 5 anterior leaflet sections; sections.
P1, P2, P3 5 posterior leaflet
accurate for both native and prosthetic mitral regurgitant defects when compared with the surgical findings. An improved understanding of the underlying pathophysiology of mitral regurgitation has led to advances in surgical therapies [12, 13]. Specifically, in selected patients, surgical repair has become the treatment of choice [14]. Indeed, valve repair is a particularly attractive option, as it avoids the morbidity associated with both valvular prostheses and long-term anticoagulation [14]. The mitral pathology most amenable to surgical repair involves disease localized to the posterior leaflet or to a focal portion of the anterior leaflet. Conversely, repair is less often successful when there is extensive disease of the anterior leaflet or when there is incomplete closure of the mitral valve, and in such cases, valve replacement is often required [1, 15, 16]. Advances in echocardiographic technology have improved the detail in which the mitral leaflets can be examined in real time. Although monoplane and biplane transesophageal echocardiographic transducers can visualize select portions of the mitral leaflets, limited imaging planes make a thorough assessment of the entire valve technically demanding. Accordingly, considerable expertise is necessary to obtain adequate information regarding the exact site of mitral leaflet pathology when such transducers are used. With the newer multiplane imaging probes, the entire mitral valve can be visualized with relatively little instrument manipulation, allowing adequate imaging in almost any given anatomic or physiologic variation. Advances in both surgical techniques and imaging technology have necessitated the development of a sysTable 2. Prosthetic Mitral Regurgitant Defect Location as Determined by Transesophageal Echocardiographya Leak Location 12:00 –2:00 2:00 – 4:00 4:00 – 6:00 6:00 – 8:00 8:00 –10:00 10:00 –12:00 a
Numbers in parentheses are percentages.
No. of Leaks (n 5 21) 3 (14) 4 (19) 4 (19) 4 (19) 2 (10) 4 (19)
tematic transesophageal echocardiographic approach for the evaluation of mitral regurgitation. To be clinically useful, valvular abnormalities must be accurately identified and their location defined relative to fixed anatomic landmarks. Several prior studies [4, 6, 8 –10, 17–22] have demonstrated the ability of transesophageal echocardiography to evaluate the mitral valve and identify its pathology. Fehske and colleagues [9] described the mitral pathologic findings in a group of patients in whom a complete examination of the leaflets was facilitated by the use of multiplane transesophageal echocardiography. Similarly, Stewart and co-workers [10] presented the advantages of multiplane transesophageal echocardiography in the complete examination of the mitral leaflets. Although these studies describe the mitral pathologic findings, they provide little information as to how one should systematically evaluate the valve. A natural instinct for the echocardiographer is to concentrate on conspicuous pathologic findings; however, the presence of leaflet sections with minor disease and the absence of pathologic findings are equally important, thus necessitating a complete and systematic examination. The current study proposes such a systematic strategy using multiplane transesophageal echocardiography and validates its efficacy against a reference standard. The distribution of defects of native mitral leaflets observed in this study was a function of the patients referred for mitral valve procedures. In our series, 70% of the lesions were localized to the posterior leaflet, a finding similar to that in a series of mitral valve repair for mitral regurgitation caused by degenerative disease [16]. In the current study, there were a small number of disagreements between the transesophageal echocardiographic and surgical findings in the native valve group. The majority of these disagreements were a consequence of localization of pathology to adjacent mitral valve sections and for the most part occurred when regurgitant lesions straddled the schematically defined sections. Similarly, in the prosthetic valve group, six of the eight disagreements were due to localization of paravalvular leaks to adjacent sections, ie, within 30 degrees between the two methods. Most of these disagreements also resulted from regurgitant jets that straddled two or more of the schematically defined sections. Of the two remaining cases of disagreement, one was an eccentrically directed paravalvular regurgitant jet that appeared echocardiographically to be a wide leak, but at operation was found to be a localized defect without extension. The surgical finding in this instance showed a focal defect with no extension. The other case was a paravalvular leak that was identified and localized preoperatively by transesophageal echocardiography but that the surgeon was unable to find once cardiopulmonary bypass was established; this paravalvular leak was still present on the postoperative transesophageal examination. This last case exemplifies the challenge surgeons have in trying to assess mitral defects in a nonphysiologic state. When the left ventricle is decompressed, the surrounding anatomy becomes distorted, making paravalvular leaks difficult to detect or localize. For native mitral
Ann Thorac Surg 1998;65:1025–31
valves, the surgeon is generally limited to assessment of the left atrial surface of the leaflets while the heart is decompressed, and this again leads to distortion of the mitral anatomy. Although insufflation of the left ventricle is a helpful technique in the evaluation of valvular and paravalvular competence, it remains a crude substitute for the true dynamic physiologic state. In contrast to the surgical setting, transesophageal echocardiography is able to evaluate the mitral valve in a normal dynamic physiologic situation without distortion of the surrounding cardiac structures. Thus, transesophageal echocardiographic findings more clearly reflect true valvular function. One last advantage of the strategy described in this study is that it provides a standardized nomenclature to localize mitral valve pathology and that this has the potential to improve communication between the surgeon and the echocardiographer. Moreover, this method should prove useful in defining the pathologic features of the mitral leaflets during preoperative transesophageal echocardiographic evaluation. This knowledge should allow decisions regarding mitral repair versus replacement to be made outside the operating room setting. The current study was limited in its ability to evaluate and compare prospectively the transesophageal echocardiographic and surgical findings without the introduction of bias. To limit this bias, the transesophageal echocardiographic studies were independently reviewed off line before comparison with the surgical findings as outlined in the detailed surgical report. Nonetheless, bias may still have been introduced during the performance of the intraoperative transesophageal echocardiogram, at which time cursory echocardiographic findings were relayed to the surgeon that could potentially have influenced the reporting of mitral pathology. A promising approach to localizing mitral pathology is through three-dimensional echocardiographic reconstruction of regurgitant defects and their associated color Doppler jets, but at present, this procedure is both time-consuming and incapable of giving “on-line” information [23]. The approach used in the current study provides a rapid and accurate method for the general echocardiographer to use until the time that on-line three-dimensional mitral valve reconstruction becomes available for surgical decision making. In conclusion, this transesophageal echocardiographic strategy provides a systematic method to accurately localize native and prosthetic mitral regurgitant lesions. By accurately localizing these defects, this strategy has the potential to improve the preoperative assessment, communication, and decisions about patients with significant mitral regurgitation.
References 1. Cosgrove DM, Stewart WJ. Mitral valvuloplasty. Curr Probl Cardiol 1989;14:359 – 415. 2. Dion R. Ischemic mitral regurgitation: when and how should it be corrected? J Heart Valve Dis 1993;2:536– 43. 3. Odell JA, Orszulak TA. Surgical repair and reconstruction of valvular lesions. Curr Opin Cardiol 1995;10:135– 43. 4. De Simone R, Lange R, Saggau W, Gams E, Tanzeem A, Hagl
FOSTER ET AL MITRAL DEFECTS BY TEE
5.
6.
7.
8.
9. 10. 11. 12.
13.
14. 15. 16. 17.
18.
19.
20.
21. 22. 23.
1031
S. Intraoperative transesophageal echocardiography for the evaluation of mitral, aortic and tricuspid valve repair. A tool to optimize surgical outcome. Eur J Cardiothorac Surg 1992; 6:665–73. Himelman RB, Kusumoto F, Oken K, et al. The flail mitral valve: echocardiographic findings by precordial and transesophageal imaging and Doppler color flow mapping. J Am Coll Cardiol 1991;17:272–9. Freeman WK, Schaff HV, Khandheria BK, et al. Intraoperative evaluation of mitral valve regurgitation and repair by transesophageal echocardiography: incidence and significance of systolic anterior motion. J Am Coll Cardiol 1992;20: 599 – 609. Homma S, Mullis S, Di Tullio MD, Schneller S, Apfelbaum M, Smith C. A transesophageal echocardiographic demonstration of ischemic mitral regurgitation. J Am Soc Echocardiogr 1993;6:341– 4. Pieper EP, Hellemans IM, Hamer HP, et al. Additional value of biplane transesophageal echocardiography in assessing the genesis of mitral regurgitation and the feasibility of valve repair. Am J Cardiol 1995;75:489–93. Fehske W, Grayburn PA, Omran H, et al. Morphology of the mitral valve as displayed by multiplane transesophageal echocardiography. J Am Soc Echocardiogr 1994;7:472–9. Stewart WJ, Griffin B, Thomas JD. Multiplane transesophageal echocardiographic evaluation of mitral valve disease. Am J Cardiac Imaging 1995;9:121– 8. Kumar N, Kumar M, Duran CM. A revised terminology for recording surgical findings of the mitral valve. J Heart Valve Dis 1995;4:70–5. Seward JB, Khanderia BK, Freeman WK, et al. Multiplane transesophageal echocardiography: image orientation, examination technique, anatomic correlations, and clinical applications. Mayo Clinic Proc 1993;68:523–51. Stewart WJ, Currie PJ, Salcedo EE, et al. Evaluation of mitral leaflet motion by echocardiography and jet direction by Doppler color flow mapping to determine the mechanism of mitral regurgitation. J Am Coll Cardiol 1992;20:1353– 61. Akins CW, Hilgenberg AD, Buckley MJ, et al. Mitral valve reconstruction versus replacement for degenerative or ischemic mitral regurgitation. Ann Thorac Surg 1994;58:668–76. Antunes MJ. Mitral valve repair into the 1990s. Eur J Cardiothorac Surg 1992;6:S13– 6. David TE, Armstrong S, Sun Z, Daniel L. Late results of mitral valve repair for mitral regurgitation due to degenerative disease. Ann Thorac Surg 1993;56:7–14. Hellemans IM, Pieper EG, Ravelli ACJ, et al. Comparison of transthoracic and transesophageal echocardiography with surgical findings in mitral regurgitation. Am J Cardiol 1996; 77:728–33. Flachskampf FA, Hoffmann R, Franke A, et al. Does multiplane transesophageal echocardiography improve the assessment of prosthetic valve regurgitation? J Am Soc Echocardiogr 1995;8:70– 8. Alam M, Serwin JB, Rosman HS, Polanco GA, Sun I, Silverman NA. Transesophageal echocardiographic features of normal and disfunctioning bioprosthetic valves. Am Heart J 1991;121:1149–55. Chaudhry FA, Herrera C, DeFrino PF, Mehlman DJ, Zabalgoitia M. Pathologic and angiographic correlations of transesophageal echocardiography in prosthetic heart valve dysfunction. Am Heart J 1991;122:1057– 64. Maurer G, Siegel RJ, Czer LSC. The use of color flow mapping for intraoperative assessment of valve repair. Circulation 1991;84(Suppl 1):250– 8. Meloni L, Aru G, Abbruzzese PA, Cardu G, Martelli V, Cherchi A. Localization of mitral periprosthetic leaks by transesophageal echocardiography. Am J Cardiol 1992;69:276–9. Salustri A, Becker AE, Van Herwerden L, Vletter WB, Ten Cate FJ, Roetlandt JRT. Three-dimensional echocardiography of normal and pathologic mitral valve: a comparison with two-dimensional transesophageal echocardiography. J Am Coll Cardiol 1996;27:1502–10.