Author + information
- Received January 7, 2008
- Accepted January 28, 2008
- Published online March 1, 2008.
- Patrick O’Gara, MD, FACC⁎,
- Lissa Sugeng, MD, FACC‡,
- Roberto Lang, MD, FACC‡,
- Maurice Sarano, MD, FACC§,
- Judy Hung, MD, FACC¶,
- Subha Raman, MD, FACC∥,
- Gregory Fischer, MD†,
- Blasé Carabello, MD, FACC⁎⁎,
- David Adams, MD, FACC† and
- Mani Vannan, MBBS, FACC∥,⁎ ()
Reprint requests and correspondence:
Dr. Mani A. Vannan, The Ohio State University, 473 West 12th Avenue, DHLRI, Suite 200, Columbus, Ohio 43210.
Chronic degenerative mitral regurgitation (MR) is a complex problem, which requires an integrated assessment of etiology, pathophysiology, and severity to enable informed clinical decision-making. A multidisciplinary approach is required, with input from the clinician, imager, and surgeon. This review begins with a discussion of essential echocardiographic and surgical mitral valve (MV) anatomy, which dictates suitability for repair when indicated. The echocardiographic and Doppler principles, which underlie the quantitation of MR severity, are summarized in the next section, followed by a critical examination of left ventricular systolic function in this disorder. A brief discussion of the important role of imaging in the developing field of percutaneous MV repair is included. The methodical and objective noninvasive assessment of degenerative MR herein reviewed is intended to help guide management decisions for patients with this challenging valve lesion.
The natural history of chronic mitral regurgitation (MR) depends intimately on its etiology, the severity of the left ventricular (LV) volume overload, left and right ventricular contractile performance, and the superimposition of other destabilizing influences such as atrial fibrillation. Table 1 lists the most common causes of chronic MR with reference to the valve level(s) most commonly affected by the specific disorders. These causes are not mutually exclusive. There are important differences in the outcomes of patients with ischemic versus nonischemic MR, related both to the effects of recurrent myocardial ischemia and to the significant contribution of LV remodeling to survival and function. The approach to nonischemic MR, which will be the focus of this review, may lend itself to a more uniform set of principles focused more specifically on the valve lesion and LV function.
The management of chronic nonischemic MR poses a unique set of challenges, predicated in part on the more diverse causes of this valve disorder, the difficulty in assessing LV contractile performance in the setting of reduced afterload, the subtle and oftentimes late appearance of symptoms, the misapplication of vasodilator therapy, uncertainty regarding the indications for surgery in the asymptomatic patient, and the still-evolving adoption of operative repair techniques. The potential role of percutaneous mitral valve (MV) repair is the subject of active investigation. Recent improvements in surgical outcomes have frame-shifted the timing of operation earlier in the natural history of appropriately selected patients with chronic nonischemic MR and have forced clinicians to balance the upfront risks of surgery (death, stroke, bleeding, failure of repair) versus the delayed consequences of progressive LV volume overload (heart failure, atrial fibrillation, death) which, for some patients, may not develop for decades, if at all. To date, large-scale randomized controlled clinical trials comparing different treatment strategies have been lacking. Comparisons across available registry and single-center studies have been hampered by several factors, including varying definitions of MR severity, heterogeneous patient populations, selection bias, lack of standardized medical and surgical therapies, and dissimilar surrogate and clinical end points. Management strategies have therefore evolved empirically, and guidelines have relied more on expert consensus opinion than on objective measures of efficacy and safety. Because the data available from simple bedside assessment are inherently imprecise and because the consequences of clinical decision-making are potentially so profound, additional information is required to make appropriate treatment recommendations. Noninvasive imaging and, in particular, echocardiography, helps fill this void and plays a critical role in the initial and longitudinal assessment of patients with chronic MR. In this review, we will examine the application of standardized, noninvasive imaging to medical and surgical decision-making in patients with chronic, nonischemic MR. Critical components of the noninvasive evaluation include information regarding MV anatomy, MR severity, LV size and systolic function, and associated findings such as estimated pulmonary artery (PA) pressures. In most instances, such information can be obtained with standardized echocardiographic protocols, although cardiac magnetic resonance (CMR) and computed tomography may provide supplementary information in selected patients.
Mitral Valve Anatomy
Normal MV anatomy
The MV is a complex 3-dimensional (3D) structure comprising 5 distinct but highly integrated levels:
The MV has 2 leaflets, namely anterior (AML) and posterior (PML) leaflets (Fig. 1). The PML has a quadrangular shape and is attached to approximately three-fifths of the annular circumference. It typically has 2 well-defined indentations that divide it into 3 individual scallops identified as P1 (anterior or medial scallop), P2 (middle scallop), and P3 (posterior or lateral scallop), according to the Carpentier classification. The 3 opposing segments of the AML are designated as A1 (anterior segment), A2 (middle segment), and A3 (posterior segment). The segmental differentiation of the leaflets is an important tool to describe specific anatomic conditions. The PML indentations are believed to aid in leaflet opening during diastole. The height of the normal PML is less than one-half of the AML. However, both leaflets have similar surface areas. The AML has a semicircular shape and attaches to approximately two-fifths of the annular circumference. The fibrous continuity between the AML and the aortic annulus in the area of the left and noncoronary cusps is designated the aortic-mitral curtain. The free edge of the AML is usually continuous, without indentations. The motion of the AML defines an important boundary between the inflow (during diastole) and outflow (during systole) tracts of the LV. A rough or coaptation zone exists on the atrial surface of the peripheral leaflet margins (Fig. 2). The normal MV may have a coaptation length of several millimeters to ensure valve competency against normal end-systolic pressure.
The commissures define a distinct area where the AML and PML come together. The amount of commissural tissue varies greatly; commissures may exist as distinct leaflet segments, similar to adjacent segments of the PML. More commonly, the commissures constitute several millimeters of leaflet tissue, which provide continuity between the AML and PML at their annular insertion. Commissural chordae have a distinct configuration, providing support to the commissure as well as to adjacent leaflet segments. In Carpentier’s segmental anatomy, the anterior-lateral commissure is referred to as AC and the posterior-medial commissure as PC.
The chordae tendineae make up theleaflet suspension system that ultimately determines the position of and tension on the leaflets at end-systole. Chordae originate from the fibrous heads of the papillary muscles and may be classified according to their site of insertion on the leaflets. Marginal or “primary” chordae insert on the free margin of the leaflets, prevent marginal prolapse, and align the rough zones to ensure coaptation. Intermediate or “secondary chordae” insert on the ventricular surface of the body of the leaflets and primarily prevent billowing and reduce tension on leaflet tissue (Fig. 3). They may also play a role in dynamic ventricular shape and function, because of their contribution to ventricular-valve continuity. Basal or “tertiary chordae” connect the base of the PML and mitral annulus to the papillary muscle.
The mitral annulus constitutes the anatomical junction between the LV and the left atrium (LA) and serves as the insertion site for the 2 leaflets. It is divided segmentally into anterior and posterior components. The anterior portion of the mitral annulus attaches to the fibrous trigones. The right fibrous trigone comprises portions of the mitral, tricuspid, and aortic annuli and the membranous septum. The left fibrous trigone is an area made up of the left fibrous borders of the aortic-mitral curtain. The posterior mitral annulus is less well developed, and the fibrous skeleton of the heart is discontinuous in this region, explaining why the posterior portion of the annulus is prone to enlarge with LV or LA dilation. The mitral annulus has a 3D saddle shape. During systole, the circumference decreases as the commissural areas move apically, the aortic root bulges and the annulus contracts.
Papillary muscles and the LV
There are 2 papillary muscles rising from the area between the apical and middle third of the left ventricular free wall. The anterolateral papillary muscle is often composed of one body or head, whereas the posteromedial papillary muscle may have 2 or more heads. Each papillary muscle provides chordae to both leaflets. The axial relationship of the chordae prevents abrasion and dyssynchrony. The anterolateral papillary muscle blood supply may originate from one or more left coronary branches. The posteromedial papillary muscle has a singular blood supply (right or left circumflex depending on dominance) and is therefore particularly prone to injury from myocardial infarction. The attachment of the papillary muscles to the lateral wall of the LV means the ventricle itself is also an important part of the MV complex. Any change in ventricular geometry that affects papillary muscle position can change the axial relationship of the chordae and leaflets, resulting in poor coaptation.
Pathophysiologic MV anatomy
Carpentier et al. (1) also have proposed a pathophysiologic triad by which MR can be described. The triad consists of the disease (the underlying etiology, such as Barlow’s disease or fibroelastic deficiency), the lesions (e.g., chordal elongation or rupture, leaflet distension, annular, leaflet and/or papillary muscle calcification, and annular dilation), and the dysfunction (defined according to the systolic position of the leaflet margins in relation to the annular plane). Type I dysfunction implies normal leaflet motion with isolated annular dilation, leading to poor leaflet coaptation. Type II dysfunction is defined by excess motion of the margin of a leaflet segment above the annular plane and is the most common type of dysfunction in degenerative MV disease. Restricted motion of the leaflet margin defines Type III dysfunction, which can be further subdivided based on restriction of motion in both systole and diastole (Type IIIa—usually due to fibrosis of the subvalvular apparatus) or only in systole (Type IIIb, which usually results from ventricular remodeling with papillary displacement and leaflet tethering). Although most patients with degenerative disease present with chordal elongation or rupture with Type II dysfunction and associated annular dilation, any combination of dysfunction can occur based on the lesions.
The 2 main causes of degenerative MV disease are Barlow’s disease and fibroelastic deficiency (2,3). Barlow’s disease results in myxoid degeneration of the MV, creating excess tissue in multiple valve segments, chordal thickening and elongation, annular dilation, and a tendency to calcification. Chordal rupture is relatively uncommon. The Type II dysfunction observed in Barlow’s patients occurs in midsystole, resulting in a nonejection click and a mid- to late-systolic murmur due to chordal elongation and excess tissue. Barlow’s disease is usually diagnosed in young adulthood, and patients often remain asymptomatic with well-preserved LV size and function for many years. In contrast, fibroelastic deficiency results from loss of mechanical integrity due to abnormalities of connective tissue structure and/or function, leading to chordal thinning, elongation, and/or rupture, with classic findings of prolapse and MR of varying severity. The prolapsing segment may be distended, but the remaining segments of the valve may be entirely normal. Fibroelastic deficiency is the most common form of organic mitral valve disease for which surgery is required. The noninvasive assessment of these 3 pathophysiologic elements informs clinical decision-making and helps direct surgical referral and planning. Increasing degrees of complexity should prompt preferential referral to a specialized center of excellence.
Surgical and echocardiographic MV anatomy
The MV can be displayed in 3 different ways as shown in Figures 4A to 4C. The anatomically correct orientation displays the valve from the base of the heart as if the imager were positioned in the left atrium looking down toward the LV. In this view, the patient’s left and right side correspond to the imager’s left and right side, whereas the AML and PML appear in their appropriate positions. The lateral and medial aspects of the valve are to the left and right, respectively. The transesophageal view corresponds to the orientation of the MV found in the transgastric basal short axis view (Fig. 4B). This anatomically “incorrect display” used in all ultrasound platforms results from rotating the anatomical view 180°. In this view, the operator’s left and right sides are reversed relative to the patient. The surgeon’s view (Fig. 4C), which can be easily obtained with real-time transesophageal echocardiography (TEE), displays the valve in a manner similar to the one encountered when examining it through an opened LA, standing to the right of the patient.
A systematic 2D TEE examination of the MV consists of 4 standard midesophageal views (4-chamber, bicommissural, 2-chamber, and long-axis views) and the transgastric basal short-axis view (Fig. 2). It should be noted that the classification of the MV scallops in any given plane may vary according to the individual anatomy. It is crucial to ensure that the imaging plane does not fore-shorten the LV because this may lead to misidentification of the individual scallops. Accordingly, orientation to internal landmarks such as the commissures is paramount to enhance diagnostic accuracy. This systematic approach is also useful in identifying the incompetent segment. A regurgitant jet arising from the left coaptation point indicates involvement of P3/A3, whereas a jet arising from the right suggests involvement of A1/P1. With the use of 3D TEE, the MV annulus and leaflets are best displayed when obtained in zoom mode to avoid stitch artifacts that may occur in a wide-angled acquisition. To simulate a surgeon’s view of the valve, the 3D TEE image is positioned with the aortic valve at the 11-o’clock position (Fig. 1). The LV view is another orientation from which to assess the valve, particularly in patients with pathology involving the submitral apparatus. If the subvalvular mitral apparatus or LV constitutes the area of interest, a wide-angled acquisition consisting of 4 smaller wedges of volume obtained over 4 cardiac cycles should be obtained to demonstrate papillary muscle location or quantitate LV function. Figures 5 and 6⇓⇓ demonstrate examples of patients with mitral valve flail and prolapse as visualized with 2D TEE and 3D TEE.
The 3D TEE parameters of interest include: 1) the major 3D axes of the annulus, anteroposterior and anterolateral-posteromedial diameters, and annular height; 2) 3D curvilinear leaflet lengths of the anterior middle scallop, across its central portion from the annulus to the central coaptation border and of the corresponding posterior middle scallop; 3) anterior and posterior leaflet surface areas, excluding the leaflet contact areas; 4) the angle between the aortic valve annulus (assumed to be planar) and a least-squares fit plane of the mitral annulus; and 5) the distances between the commissures and papillary muscles (Fig. 7). Measurement of these parameters may aid in understanding the mechanism of MR, improve surgical planning, and delineate annular and annuloplasty ring dynamics before and after operation. Higher-resolution 3D TEE images enable improved quantitation of LV function.
Quantitation of MR
Quantitation of MR severity is essential for clinical decision-making. Referral for surgery is not appropriate for the asymptomatic patient without confirmation that the MR is severe by standardized criteria. The use of descriptive terms such as “moderately severe” can be highly subjective and influenced by any one of several treatment biases. The importance of MR quantitation has been emphasized by the guideline writing committees of the American College of Cardiology/American Heart Association and the European Society of Cardiology (4). The severity of MR, in turn, should dictate the associated degree of LV and LA enlargement, as well as PA pressure elevation, in patients with isolated valve disease. The use of 2D Doppler echocardiographic quantitation proves sufficient for most patients, though 3D TEE and CMR may provide more accurate information in selected cases.
A comprehensive approach to the assessment of MR severity is required. Specific findings related to MV anatomy, supportive findings, which include chamber sizes, intra-cardiac flows, and PA pressures, as well as quantitative findings regarding regurgitant volume (RVol) and effective regurgitant orifice area (ERO), constitute the basic features of an integrated assessment. Reporting of MR severity should be consistent with the ASE criteria for descriptive and semi-quantitative grading shown in Table 2.
Qualitative echo-Doppler assessment
The entire MV apparatus should be examined carefully and the color scale optimized. Mild MR is characterized qualitatively as a small jet confined to early or late systole with small/absent flow convergence and a narrow vena contracta. In the absence of intrinsic LV systolic or diastolic dysfunction, the LV and the LA dimensions also are normal. Qualitative assessment of larger or more eccentric jets is more difficult. Flail leaflet or ruptured papillary muscle most often is associated with severe MR, although the regurgitant load may be only moderate by quantitative criteria in as many as 15% of patients. Systolic flow reversal in the pulmonary veins by pulsed-wave Doppler is relatively easy to demonstrate and is specific for severe MR. Other signs considered specific for severe MR are the presence of a large flow convergence, a jet with a wide vena contracta, and a large, wall-impinging or swirling jet pattern. Interpretation of these signs is subject to wide interobserver variability. High jet density by continuous-wave Doppler, a dominant mitral E, and dilated LA and LV chambers, support the diagnosis of severe MR. There are several additional limitations to the qualitative assessment of MR. First, there are no specific criteria for the designation of moderate MR, other than the absence of findings consistent with either mild or severe MR (4). Second, interpretation of color flow patterns of MR can be highly subjective, thus blurring the distinction between moderate and severe. It is widely appreciated that eccentric jets tend to underestimate and central jets to overestimate RVol (5). Third, although specific signs have high positive predictive value, they lack sensitivity for the detection of severe MR. These limitations have led to the development of quantitative methods for assessment of MR (4).
Quantitative Doppler echocardiographic assessment
Quantitation is based on hydrodynamic principles which rely on the noncompressibility of blood and the continuity or conservation of mass. Flow can be calculated as: flow = (vessel area) · (mean velocity of blood). These concepts are used to measure the 3 parameters indicative of MR severity: 1) RVol (the volume in ml/beat regurgitated each systole), a measure of absolute volume overload; 2) regurgitant fraction (RF, the percentage of the total LV stroke volume represented by the RVol), a measure of relative volume overload; and 3) ERO (the mean area of the systolic regurgitant orifice), a measure of lesion severity.
These parameters of MR severity are measured with the use of 3 validated methods: 1) quantitative Doppler (4) is based on measurement by pulsed-wave Doppler of mitral and aortic stroke volumes using the product of mitral and aortic annular areas times their integral of forward blood velocity (tissue velocity integral [TVI]). Regurgitant volume is calculated as the difference between mitral and aortic stroke volume, RF as the ratio of RVol to mitral (total) stroke volume, and ERO as the ratio RVol to the regurgitant jet TVI; 2) quantitative 2D echocardiography (4) is based on the same principles, but mitral stroke volume is replaced by that of LV stroke volume by tracing end-diastolic and end-systolic LV volumes; and 3) proximal isovelocity surface area (PISA), which focuses on the flow convergence proximal to the regurgitant orifice (Fig. 8), as observed with color-flow imaging (6). The color scale baseline is shifted downward to ensure measurability of the flow convergence radius and velocity. Flow through the convergence zone (and hence through the regurgitant orifice) is calculated as the product (area of flow-convergence hemisphere) · (aliasing velocity). This velocity term is indicated by the machine (junction blue-yellow). Effective regurgitant orifice area is calculated as the ratio of flow to peak regurgitant velocity and RVol as the product of (ERO) · (regurgitant TVI) (7). Thus, it is possible during a single examination to measure RVol and ERO by the use of multiple methods.
Accuracy and reproducibility
Quantitation requires attention to detail, repetition, and quality control. The PISA method requires close zooming and downshift of the color baseline tailored to obtain an unconstrained flow convergence of almost circular shape (Fig. 9). The flow convergence radius should be measured at peak regurgitant velocity, which usually occurs near the T-wave on the electrocardiogram. Regurgitant flow, and therefore ERO, may vary throughout systole as a function of the cause of the MR. Continuous-wave Doppler should be recorded with a single-crystal transducer, and attention paid to avoid incorporation of partial systolic signals (early and late). Simplified estimates of RVol may be derived from the ratio of flow/3.25. Multiple measurements of these parameters should be made and the values for RVol and ERO averaged.
Emerging technologies for the assessment of MR severity may prove useful. Real-time 3D TEE holds promise for volume measurement, jet characterization, and definition of proximal flow convergence. Cardiac magnetic resonance is of limited value for assessment of MV structure and motion, but it can be use to provide highly accurate quantitation of LV and regurgitant volumes. The ERO measurements cannot be reliably obtained with this technique.
Interpretation and integration
The qualitative and quantitative characterization of MR must be integrated with clinical information to determine optimal treatment and long-term follow-up. For patients with isolated mild or moderate nonischemic MR with preserved LV size and function, expectant management with periodic surveillance is appropriate. It is for patients with isolated, severe MR that timing of surgery becomes critical. Many such patients are asymptomatic and thus reliance on objective and standardized noninvasive data is necessary for clinical decision-making. Designation of MR as severe must be verifiable across the spectrum of parameters reviewed above and summarized in Table 2.
The LV in Chronic MR
Figure 10 depicts the stages that an untreated patient with severe MR might experience (8). In acute, severe MR (Fig. 10B), there is sudden distension of the LA, producing LA hypertension and pulmonary congestion. Increased LV preload is manifested by slightly increased end diastolic volume (EDV) whereas the regurgitant pathway for ejection into the LA decreases LV afterload and end systolic volume (ESV). These changes act in concert to increase total stroke volume (SV) but not enough to normalize forward SV. Thus, the patient has decreased forward output and pulmonary congestion and is in heart failure despite normal LV contractile function. If the patient survives the acute insult and is left untreated, or if the MR develops more gradually over time, he or she may enter a chronic compensated phase (Fig. 10C). In this phase, eccentric hypertrophy has developed, increasing EDV. The radius term in the Laplace equation is increased, but systolic wall stress does not increase. The increase in EDV more than offsets the increase in ESV, so that total SV is increased, in turn increasing forward SV. An enlarged, compliant LA can accommodate the RVol at lower pressure, relieving pulmonary congestion. In this phase, the patient may be completely asymptomatic. How long the patient remains in this phase is quite variable, depending on whether the MR worsens and probably on genetic variations in tolerance of the volume overload. Eventually, severe MR leads to muscle dysfunction and decompensation (Fig. 10D). In this phase, a weakened LV can no longer shorten adequately and ESV increases, decreasing total and forward SV while increasing LA and LV filling pressures. Symptoms of heart failure may develop, but some patients remain asymptomatic or fail to recognize deterioration in functional status. Importantly, ejection fraction (EF) may remain in the “normal” range, belying the presence of LV dysfunction (9).
Ejection fraction has, during the last 4 decades, become the most widely used descriptor of LV function. All imaging techniques can be used to determine LVEF. However, the issue at hand for the clinician is to determine whether or not the patient with MR is entering the phase of early LV muscle dysfunction, a phase when correction of MR must be undertaken to prevent permanent myocardial damage. The major function of any muscle including the myocardium is to develop force. Obviously, EF has no force term in its expression. What the clinician would like to know is whether contractility, the innate ability of the myocardium to generate force, has begun to decline. The ideal measure of contractility would be a measure independent of afterload and preload, sensitive to changes in inotropic state, insensitive to LV size, and easy to apply (10). Unfortunately, no such measure exists but EF, which is easy to apply, has become the cornerstone of LV function assessment.
Unfortunately, EF is especially suspect in patients with MR because it is load sensitive and both afterload and preload are altered in patients with MR (11). Increased pre-load and reduced afterload act in concert to increase EF beyond what would be expected from inotropic state. Thus, patients with substantial LV dysfunction may still have a normal EF (9,12) To overcome this deficiency the clinical use of EF in MR patients, a greater limit for normal has been set. Although in most subjects the lower limit of normal for EF is 0.55, MR patients with EF <0.60 have been noted to have an impaired prognosis (9,13). The EF is further compromised because it is often estimated rather than measured precisely, making it unclear whether the distinction between an EF of 0.60 and 0.55 is accurate. Because of the imprecision of EF in assessing LV function, especially in MR, other measures have been explored.
End-systolic dimension (ESD) and volume
A number of studies have examined whether ESV or ESD is predictive of outcome in chronic MR (14–16). The volume to which the LV contracts at the end of systole is determined by contractility, afterload, and eccentric remodeling, but not by pre-load. Thus, ESV and ESD are independent of one of the factors confounding the use of EF in assessing LV function in MR. Additionally, afterload tends to increase in the late stages of MR whereas, at the same time, contractility tends to decrease. An increased LV-ESD (>40 mm) and an LVEF <0.60 are indicators of LV systolic dysfunction, portend poor long-term prognosis, and generally are accepted as indications for surgery even in the absence of symptoms (17). Thus, the determination of ESV or ESD is an important component of the noninvasive assessment of patients with MR. Left ventricular dimensions can be accurately obtained with 2D techniques. Volume determinations are more precise with 3D TEE and CMR.
The response to exercise stress assessed with echocardiography has also been used to identify subclinical LV systolic dysfunction in patients with asymptomatic MR. Impaired contractile reserve with exercise (defined as an increase in EF <0.04) has been shown to predict worsening LV function in medically treated patients. Reduced reserve also helps to predict immediate post-operative LV systolic dysfunction and early cardiac events in patients undergoing surgery (18).
Doppler imaging of LV myocardial segments has been used as a surrogate marker of LV function (19). Systolic tissue Doppler velocities have been shown to correlate with LV function, to identify subclinical LV dysfunction, and to predict post-operative LV function in patients with asymptomatic MR (20,21).
Myocardial strain is defined as a change in length compared with a reference initial length. In contrast to velocity of contraction, strain imaging is not influenced by global cardiac motion but may be influenced by large variations in loading conditions (22). Strain rate is the change of strain over time (23). Both strain and strain rate have been demonstrated to correlate well with LV function, wherein strain appears to correlate with the stroke volume and strain rate with LV contractility assessed by invasively derived dP/dt (24). Strain and strain rate help identify subclinical LV dysfunction in patients with asymptomatic severe MR and correlate with contractile reserve with exercise; strain and strain rate are significantly greater in patients with adequate contractile reserve (25). Moreover, strain and strain rates have been shown to decrease even before changes in LV chamber dimensions occur, i.e., before LV systolic dimension exceeds 4.5 cm (26). Limitations of strain imaging include its dependence on pre-load, afterload, and imaging angle. To correct for the influence of preload on strain/strain rate and obtain a more pure measure of contractility, Marciniak et al. (26) devised geometry-compensated deformation indices, which are calculated by dividing strain and strain rate by EDV. The geometry compensated deformation indices are decreased in patients with LV systolic dimension >4.5 cm (22). However, the incremental value of accurately identifying preclinical LV dysfunction remains to be determined.
Role of Imaging in Ongoing Investigations of Percutaneous MV Repair
Recently, there has been considerable interest in percutaneous MV repair. One approach, which adopts the principles of the Alfieri surgical technique, involves the deployment of a clip or stitch to bring the leaflet edges together (Fig. 11). In the process, a double orifice valve is created. A second approach aims to reduce mitral annular circumference via placement of a device in the coronary sinus (Fig. 11). These coronary sinus devices reshape the posterior mitral annulus.
Pre-, intra-, and post-procedural imaging is critical to the successful use of these devices. There are 4 basic imaging objectives: 1) evaluate the mechanism and severity of MR; 2) determine anatomic suitability for device repair; 3) guide deployment of the device; and 4) assess stability and outcome of repair. Table 3 lists key imaging factors for each of the percutaneous techniques. For edge-to-edge repair, TEE is the most widely used approach to identify the location of abnormal anatomy, and guide the procedure. Mid-esophageal long-axis and commissural TEE views are particularly helpful to center the clips over the MV orifice, and the transgastric short-axis view is used to grasp the leaflet edges. A successful and smooth procedure is highly dependent on communication between the echocardiographer and interventionalist. Guidance with TEE may be significantly improved with the use of 3D imaging, allowing for more precise delivery of the catheter towards the leaflet edges.
Coronary sinus (CS) technique
These devices mimic surgical annuloplasty, which reduces mitral annular area by moving the posterior annulus anteriorly or by decreasing the circumference of the posterior annulus. In developing nonsurgical approaches to MV repair, it has been appreciated that the cardiac veins typically encircle the mitral annulus. These veins could serve as the anatomic target for percutaneously deployed devices that effectively reduce the ERO by exerting inward pressure on the mitral annulus. The typical configuration of the major cardiac veins is shown in Figure 12. The CS and great cardiac vein (GCV) form a continuous cardiac venous landing zone for a MV repair device delivered transvenously. However, cardiac veins exhibit great variability in extent, distribution, and anatomic location. The first variable to consider for successful device deployment is the anatomy of the CS-GCV. Distance of the CS-GCV from the mitral annulus in the superior-inferior dimension as well as in the radial direction from the center of the mitral valve may be important (27). This variance may be an important factor determining the likelihood of procedural success. A device that encircles primarily the LA free wall may simulate cor triatriatum without affecting a decrease in the MV ERO. Other relevant variables include the length of the CS-GCV and the extent to which these veins overlie the mitral annulus. Another important anatomic variable is the course of the left circumflex artery and its branches (28).
Coronary venous imaging may be preferred by the interventionalist for immediate recognition of anatomic landmarks, but does not provide concomitant information regarding cardiac and mitral apparatus anatomy that may ultimately determine procedural success. Simultaneous transesophageal or intracardiac echocardiography help overcome these deficiencies. The use of cardiac computed tomography allows the simultaneous visualization of venous, coronary artery, and cardiac anatomy. It remains to be determined what geometries pose the greatest risk for myocardial ischemia from extrinsic coronary compression. Simultaneous rendering of venous and coronary artery anatomy with cardiac computed tomography affords such definition. The use of CMR has a limited role at the present time, although recent developments in technology may improve the utility of this technique (29). A multimodality imaging approach holds promise for anatomic definition of the coronary sinus and LV during the procedure. Left ventricular volumes obtained by 3D echocardiography can be superimposed on reconstructed images of the coronary sinus using high speed rotational coronary venous angiography (29). This strategy of integrating echocardiographic and angiographic data to reconstruct coronary sinus and LV anatomy holds great potential.
How should the clinician use the information gathered from the history, examination, and noninvasive imaging to construct the most appropriate management plan for any individual patient with chronic, nonischemic MR? After verification of the diagnosis, one approach begins with designation of symptom status (10). The onset of symptoms is a Class I indication for surgical referral, preferably with repair, provided: 1) the MR is severe by quantitative measures and the likely cause of the symptoms reported and 2) and LV systolic function is normal or only moderately impaired (EF >0.30, ESD <5.5 cm). For patients with more severe degrees of LV dysfunction (EF <0.30, ESD >5.5 cm), surgery is appropriate when chordal preservation (with either repair or replacement) is likely (Class IIa).
There is a relatively greater reliance on noninvasive imaging for the management of asymptomatic patients with nonischemic MR, particularly when considering the indications for surgical repair. Under these circumstances, the risk-benefit analysis must be extremely favorable and driven in large measure by the availability of an expert surgeon with a proven track record of outstanding results. Valve replacement is an unacceptable outcome for the asymptomatic patient. The following questions must be answered from the noninvasive data:
1. Is the MR severe?
a. No. Establish schedule for clinical and TTE surveillance.
b. Yes. Establish schedule for clinical and TTE surveillance and consider whether early surgery might be warranted depending on the answers to the questions which follow.
2. Is LV function preserved?
a. No. If MR is severe and EF <0.60 or ESD >4.0 cm, refer to surgery (Class I).
b. Yes. If MR is severe, and EF >0.60 with ESD <4.0 cm, consider early surgical referral depending on likelihood of repair and availability of surgical expertise (Class IIa).
3. Is PA hypertension present?
a. No. If PA systolic pressure <50 mm Hg at rest and/or <60 mm Hg with exercise, establish schedule for clinical and TTE surveillance.
b. Yes. If MR is severe and PA systolic pressure >50 mm Hg at rest or >60 mm Hg with exercise, consider early surgical referral (Class IIa).
4. What is the mechanism of the MR and therefore the feasibility of primary repair?
a. Myxomatous degeneration/fibroelastic deficiency
b. Other: Barlow’s, congenital, and so on
5. Is there a very high (>95%) likelihood of successful repair?
a. Yes. Consider early referral to an experienced surgeon at a high volume center of excellence (Class IIa).
b. No. Establish clinical and TTE surveillance.
Decision-making of this nature requires the collaborative input of the clinician, echocardiographer/imager, and the surgeon. With appropriate training, attention to detail, informed understanding of natural history, more widespread adoption of standardized evaluation and treatment protocols, and further advances in surgical and intraoperative management, outcomes for patients with nonischemic MR will continue to improve. The potential role of percutaneous repair has not yet been established.
The authors acknowledge the contributions Francesco Grigioni, MD, of Mayo Clinic, Rochester, Minnesota; and Kumudha Ramasubbu, MD, of Baylor College of Medicine, Houston, Texas, for their valuable contributions to the manuscript.
- Abbreviations and acronyms
- anterior mitral leaflet
- cardiac magnetic resonance
- coronary sinue
- end-diastolic volume
- ejection fraction
- effective regurgitant orifice area
- end-systolic dimension
- end-systolic volume
- great cardiac vein
- left atrium
- left ventricular
- mitral regurgitation
- mitral valve
- proximal isovelocity surface area
- posterior mitral leaflet
- regurgitant fraction
- regurgitant volume
- stroke volume
- transesophageal echocardiography
- Received January 7, 2008.
- Accepted January 28, 2008.
- American College of Cardiology Foundation
- Zoghbi W.A.,
- Enriquez-Sarano M.,
- Foster E.,
- et al.
- Enriquez-Sarano M.,
- Tajik A.,
- Bailey K.,
- Seward J.
- Vandervoort P.,
- Rivera J.,
- Mele D.,
- et al.
- Enriquez-Sarano M.,
- Miller F.A.J.,
- Hayes S.N.,
- Bailey K.R.,
- Tajik A.J.,
- Seward J.B.
- Carabello B.A.
- Enriquez-Sarano M.,
- Tajik A.J.,
- Schaff H.V.,
- Orszulak T.A.,
- Bailey K.R.,
- Frye R.L.
- Carabello B.A.
- Wisenbaugh T.,
- Spann J.F.,
- Carabello B.A.
- Matsumura T.,
- Ohtaki E.,
- Tanaka K.,
- et al.
- Zile Mr,
- Gaasch W.H.,
- Carroll J.D.,
- Levine H.F.
- Crawford M.H.,
- Souchek J.,
- Oprian C.A.,
- et al.
- Bonow R.O.,
- Carabello B.A.,
- Chatterjee K.,
- et al.,
- American College of Cardiology/American Heart Association Task Force on Practice Guidelines,
- Society of Cardiovascular Anesthesiologists,
- Society for Cardiovascular Angiography and Interventions,
- Society of Thoracic Surgeons
- Leung D.Y.,
- Griffin B.P.,
- Stewart W.J.,
- et al.
- Agricola E.,
- Galderisi M.,
- Oppizzi M.,
- et al.
- Nixon J.V.,
- Murray R.G.,
- Leonard P.D.,
- Mitchell J.H.,
- Blomqvist C.G.
- D’hooge J.,
- Heimdal A.,
- Jamal F.,
- et al.
- Greenberg N.L.,
- Firstenberg M.S.,
- Castro P.L.,
- et al.
- Marciniak A.,
- Claus P.,
- Sutherland G.R.,
- et al.
- Tops L.F.,
- Van de Veire N.R.,
- Schuijf J.D.,
- et al.
- Maselli D.,
- Guarracino F.,
- Chiaramonti F.,
- et al.