Physical Address
304 North Cardinal St.
Dorchester Center, MA 02124
The complexity of the mitral valve is such that two-dimensional (2D) imaging does not adequately describe its anatomy and function. Indeed, the mitral valve apparatus is a dynamic three-dimensional (3D) structure composed of the saddle-shaped annulus; two asymmetric leaflets; multiple chordae tendineae of various lengths, thicknesses, and points of attachment; the left ventricular wall and the attached papillary muscles; and parts of the left atrium ( Figure 5-1 ). During normal function, this array of parts constantly shifts in a complex but defined pattern. Normal alignment of all aspects of this biologic machine is required to avoid dysfunction. Therefore, full understanding of the mitral valve anatomy is not ascertained without 3D imaging. Real-time 3D echocardiography (RT3DE) gives echocardiographers a rapid and easily accessible method of identifying the entire mitral valve apparatus in most patients. As a prominent structure in the posterior left heart, the mitral valve often can be displayed in 3D using either a transthoracic approach or a transesophageal approach. Furthermore, detailed volumetric and positional analysis of the various mitral valve components often is possible ( Table 5-1 ). RT3DE, however, is a novel modality, and its use in mastering the analysis of the intricate mitral valve requires a substantial time commitment. This chapter highlights the use of RT3DE in imaging the normal mitral valve, demonstrating usual transthoracic and transesophageal views for each of its components. It also provides an overview of typical 3D quantitative analyses of the size and position of the mitral valve structures.
Key Measurements | Normal Values | Reference |
---|---|---|
Mitral Valve Annulus | ||
Anteroposterior dimension | 30.8 ± 4.4 mm | Sonne et al |
24 ± 1 mm | Kwan et al | |
Commissural dimension | 35.1 ± 4.9 mm | Sonne et al |
28 ± 1 mm | Kwan et al | |
Circumference | 10.5 ± 1.4 mm | Sonne et al |
10.0 ± 0.8 mm | Watanabe et al | |
Height | 4.3 ± 2.1 mm | Sonne et al |
3.0 ± 1.0 mm | Flachskampf et al | |
4.5 ± 1.1 mm | Watanabe et al | |
Area/BSA (end diastole) | 4.6 ± 1.3 cm 2 /m 2 | Qin et al |
5.9 ± 1.2 cm 2 /m 2 | Flachskampf et al | |
Area/BSA (mid systole) | 5.1 ± 0.8 cm 2 /m 2 | Sonne et al |
Area/BSA (end systole) | 3.8 ± 1.1 cm 2 /m 2 | Qin et al |
Area change | 23.8% ± 5.1% | Flachskampf et al |
15% ± 16% | Qin et al | |
Linear systolic motion | 16 ± 3 mm | Qin et al |
10 ± 3 mm | Flachskampf et al | |
Mitral Valve Leaflets | ||
Volume | 4.5 ± 0.7 cm 3 | Limbu et al |
Maximum tenting height | 5.3 ± 2.4 mm | Sonne et al |
3.1 ± 1.2 mm | Watanabe et al | |
5 ± 0.2 mm | Kalyanasundaram et al | |
Mean tenting height | 1.9 ± 1.5 mm | Sonne et al |
0.7 ± 0.5 mm | Watanabe et al | |
Tenting volume | 1.5 ± 0.9 cm 3 | Sonne et al |
0.45 ± 0.29 cm 3 | Watanabe et al | |
Papillary Muscles (Indexed to BSA) | ||
AL papillary to annular distance | 21.0 ± 5.8 mm/m 2 | Sonne et al |
PM papillary to annular distance | 22.3 ± 5.6 mm/m 2 | Sonne et al |
Interpapillary muscle angle | 17.6 ± 9.1 degrees/m 2 | Sonne et al |
Interpapillary distance | 10.5 ± 3.3 mm/m 2 | Sonne et al |
Left Ventricle | ||
End-systolic volume | 43.7 ± 10.7 mL | Chukwu et al |
59 ± 18 mL | Corsi et al | |
43 ± 18 mL | Zeidan et al | |
47 ± 6 mL | Nosir et al | |
End-diastolic volume | 115 ± 22.6 mL | Chukwu et al |
143 ± 30 mL | Corsi et al | |
108 ± 32 mL | Zeidan et al | |
113 ± 16 mL | Nosir et al | |
Left Atrium | ||
End-diastolic volume | 48.9 ± 25.1 cm 3 | Azar et al |
End-systolic volume | 62.3 ± 17.2 cm 3 | Azar et al |
3DE has increased the understanding of the anatomy of the mitral valve annulus more than that of any other structure in the heart. 3D reconstructions identified the mitral valve annulus as a hyperbolic paraboloid or a saddle-shaped structure with curved planes parallel to the anteroposterior axis opening upward and orthogonal curved planes that open downward ( Figure 5-2 ). This complex shape is not appreciated on 2D imaging, which tends to simplify the annulus as a planar ring. In fact, the anterior and posterior portions of the mitral valve annulus are about 5 mm higher than the medial and lateral commissural points. This height, along with the commissural diameter (the distance between the two low points) and the anteroposterior diameter (the distance between the two high points), can be easily measured with 3D imaging. The nonplanar shape of the mitral valve annulus is very important in reducing leaflet stress, which is increased when the annulus flattens. As such, the hyperbolic paraboloid annular shape with a height/commissural width ratio of 15% to 20% is preserved in many mammalian species, including humans.
RT3DE is fundamental to clarifying annular geometry. In fact, obtaining accurate annular dimensions is best done with 3D datasets because 2DE underestimates these values by 6% to 14%. The mitral valve annulus is best seen using the large-sector (3D full volume) and wide-sector focused (3D zoom) formats. In particular, computer remodeling from large-sector datasets can quickly measure and display these changes without requiring the time-consuming reconstructions from multiple 2D images. 3D volume data can be obtained from either transthoracic or transesophageal exams. These models are obtained offline by tracing the mitral valve annulus contours in relation to other structures (such as the papillary muscles and aortic valve) in several 3D views ( Figure 5-3 ). From these computer renditions, several annular dimensions, including circumference, anteroposterior (high-point) dimension, commissural (low-point) dimension, height, and annular area, can be measured ( Figure 5-4 ).
Recognizing the mitral valve annulus as a nonplanar structure is very important to understanding mitral valve function. For example, the complex shape of the mitral valve annulus has led to the overdiagnosis of mitral valve prolapse with 2DE. In certain 2D views, particularly the apical four-chamber view, which shows the mitral valve annulus in mediolateral directions, the leaflets appear to be on the atrial side of the mitral valve annulus, when, in fact, they are still inferior to the high anterior and posterior points. In 2DE imaging without fully realizing the nonplanarity of the mitral valve annulus, this normal anatomy can be mischaracterized as mitral valve prolapse.
RT3DE has also shown that annular area significantly depends on body size and height. The normal diastolic mitral valve area (which is a projection onto the least squares plane) is 5 to 6.5 cm 2 (indexed to body service area). The mitral valve annulus, however, is a dynamic structure whose complex motion throughout the cardiac cycle has not been fully ascertained. On the basis of early studies in 3DE from 2D renditions, it was believed that the mitral valve area decreases during systole by up to 24%, which is thought to occur as the anterior and lateral high points move closer together, increasing the height and eccentricity of the annulus. In addition to a decrease in annular area, there is a significant apical displacement of the annulus during systole, with an average annular motion of 16 ± 3 mm in normal hearts.
Recent investigations using RT3DE with volumetric reconstructions have further characterized this motion in both normal ventricles and those with cardiomyopathy. In disease states, the shape and dynamic function of the mitral valve can change dramatically. For example, in dilated cardiomyopathy, especially with functional mitral regurgitation, the annulus flattens, dilates, and becomes more circular. These annuli also demonstrate less dynamic variability. During systole, they may have less change in height and diameter, leading to a smaller change in the projected valve area. Most significantly, those with cardiomyopathy have less apical excursion during systole. In fact, the degree of annular displacement correlates well with ejection fraction; a linear motion less than 12 mm accurately identifies those with an ejection fraction less than 50%.
Like the annulus, the attached mitral valve leaflets are 3D dynamic structures. The anterior leaflet covers two thirds of the orifice and is attached to the part of anterior annulus adjacent to the aortic valve near the right fibrous trigone, aortic-mitral fibrosa, and left fibrous trigone ( Figure 5-5 ). The posterior leaflet, although smaller, covering only one third of the mitral orifice, is semicircular in shape and borders most of the annulus from one commissure to the other. Figure 5-5 shows the mitral valve leaflets and related cardiac structures as seen from the left atrium (the surgeon's view) and the left ventricle. Each leaflet is composed of three scallops. Scallops A1 and P1 are the most lateral, located near the left atrial appendage; the middle scallops, A2 and P2, are the most posterior; and A3 and P3 are the most medial.
With 3DE, all six scallops of the mitral valve leaflets can be clearly visualized from either the left atrium or the left ventricle. These views can be obtained offline from cropping 3D full-volume (large-sector) datasets, which can be obtained by either a transthoracic or a transesophageal approach ( Figure 5-6 and Video 5-1 ). During transesophageal exams, the 3D zoom (wide-sector focused) format also is helpful, especially to obtain real-time information on leaflet function during surgery or catheter procedures. In this case, the images are obtained by identifying the entire mitral valve on orthogonal biplane 2D images. The resulting 3D zoom display is rotated with the aortic valve in the 12 o'clock position to obtain the surgeon's view from the left atrium. In this view, normal systolic bulging of the central part of the anterior leaflet can be visualized. When this image is flipped 180 degrees, the leaflet scallops are displayed as viewed from the left ventricle ( Figure 5-7 ). In transthoracic exams, similarly detailed 3D images also can be obtained. Figure 5-8 demonstrates how clearly the mitral valve leaflets can be displayed from a 3D full-volume dataset acquired during a transthoracic exam. In this case, the image is rotated and the vertical axis is cropped from the atrial side to obtain the surgeon's view. By flipping the image 180 degrees and cropping from the ventricular side, the valve leaflets can be seen as viewed from the apex in relation to the commissures and papillary muscles.
The commissures of the mitral valve are easily identified by RT3DE. The anterolateral commissure is located next to the left atrial appendage and the left fibrous trigone and represents the point of fusion between A1 and P1. The posteromedial commissure is located next to the right fibrous trigone and represents the point of fusion between A3 and P3. A short-axis view of the mitral valve using the narrow-sector format can clearly display both commissures from the atrium and the ventricle by either the transesophageal or the transthoracic approach (see Figures 5-7 and 5-8 ). In the transesophageal exam, the mid-esophageal view allows ideal visualization of both commissures. The anatomy of the commissures is of particular importance in patients with mitral valve stenosis before balloon valvuloplasty. In fact, commissural splitting is believed to be the primary mechanism of increasing valve area during balloon valvuloplasty. Thus, the degree of noncalcified commissural fusion, as assessed by RT3DE, correlates with successful valvuloplasty and may be a better predictor of procedural success compared with traditional 2D scoring algorithms. Heavily calcified commissures are unlikely to split during the procedure and should be identified before referral for valvuloplasty.
In the evaluation of mitral stenosis, RT3DE can improve assessments of leaflet opening. In 2D, calculating valve area by planimetry is unreliable because it is difficult to determine if the traced plane intersects the valve at its smallest dimensions. This limitation, however, is overcome with 3DE imaging because the entire valve can be imaged and the short-axis plane can be cropped at the smallest area. As demonstrated in several studies, this method may be superior to estimating true mitral valve area when compared with 2D measurements by pressure half-time, proximal isovelocity surface area, and planimetry. Some have suggested that 3D planimetry may be more accurate than even invasive measurements of mitral valve area.
With RT3DE, it is also possible to ascertain mitral valve leaflet volume. As rheumatic disease progresses, the leaflets thicken. The degree of thickening correlates with complications such as atrial fibrillation and death. Leaflet volume also can be assessed with RT3DE, which may have as much prognostic value as mitral valve area.
RT3DE allows full assessment of the mitral valve leaflet motion in the majority of cases. Even during transthoracic exams, all three scallops of the anterior and posterior leaflets are completely visualized in up to 84% and 77% of cases, respectively. This technology is of particular use in identifying the exact location of leaflet pathology. On transesophageal echocardiography (TEE), it is possible to see each individual scallop with 2D imaging alone, but this process is cumbersome because the entire leaflets cannot be seen in the same view. To reconstruct all six scallops, an excellent working memory of mitral valve anatomy is required, and it is necessary to rotate through the various views to identify and locate pathology. With 3D imaging, however, all six scallops, along with their relationships to other structures, can be visualized at once ( Figure 5-9 ). For example, a fibroelastoma is shown in Figure 5-10 . In the 2D transesophageal views, this tumor is easily seen on the posterior leaflet, but the 3D zoom view demonstrates its exact location on P1 very close to the anterolateral commissure. In addition, mitral valve prolapse can be easily identified in 3D with the wide-sector focused view. As seen from the left atrium, the prolapsed segment is a bright bulge. From the ventricle, it is a spoon-shaped depression. Although 2DE may diagnose the presence of flail or prolapse in general, 3DE often is needed to identify the specific scallops involved. In fact, myxomatous mitral valve disease highlights the usefulness of 3D techniques. Several studies have compared the accuracy of the various echo modalities in identifying the location of mitral valve prolapse ( Table 5-2 ). 3DE also can help visualize and quantify the location and size of the orifice regurgitant area that leads to mitral regurgitation. Color Doppler can be added to full-volume acquisitions. This increased level of visualization shows the regurgitation as a column or plane. Especially in eccentric mitral regurgitation, 3D allows visualization beyond a 2D “jet,” thus better characterizing its severity. In planning mitral valve surgery, the increased level of detail proved by RT3DE is helpful to the surgeon and may increase repair rates.
N | 2DTTE (%) | 2DTEE (%) | 3DTTE (%) | 3DTEE (%) | |
---|---|---|---|---|---|
Grewal et al | 42 | — | 90 | — | 98 |
Pepi et al | 112 | 77 | 87 | 90 | — |
Manda et al | 18 | — | 50 | — | 77 |
Muller et al | 74 | — | 86-97 | — | 97-100 * |
Fabricius et al | 42 | — | 97 | — | 91 |
Sharma et al | 39 | 82 | 96 | — | 94 |
Hirata et al | 42 | 97 | 100 | — | 100 |
Agricola et al | 59 | — | 92 | 94 | — |
Garcia-Orta et al | 54 | 67-94 | 89-100 * | — | — |
* Accuracy rates reported on each segment; entire range is given here.
In addition to prolapse, RT3DE has been helpful in quantifying leaflet tethering associated with functional mitral regurgitation. Normally, the mitral valve leaflets are nearly level with the annulus (with a mean tenting height of 1.9 ± 1.5 mm). Figure 5-11 demonstrates mitral valve tenting in a normal mitral valve. In ischemic mitral regurgitation, the mitral valve leaflets may be pulled toward the ventricle by increased tension of the chordae. In extreme cases, with bileaflet restriction, the valve can form a funnel shape, allowing free central regurgitation. Several laboratories have developed methods of quantifying leaflet tethering using datasets from RT3DE. With these models, information on the mechanism of ischemic mitral regurgitation is ascertained. For example, Kwan and colleagues showed that an inferior infarct caused asymmetric tethering on the medial side of the valve.
Become a Clinical Tree membership for Full access and enjoy Unlimited articles
If you are a member. Log in here