Chapter 29. Ischemic Mitral Regurgitation

Ischemic mitral regurgitation (IMR), often termed functional mitral regurgitation, is mitral insufficiency as a result of myocardial ischemia or infarction. By definition, the mitral valve leaflets are structurally normal in IMR. The malcoaptation of the leaflets and resultant regurgitation of the mitral valve in IMR is a result of either acute papillary muscle dysfunction or rupture, or chronic changes with left ventricular remodeling and subsequent changes in geometry.

Ischemic mitral regurgitation (IMR) is associated with decreased quality of life and long-term survival. Although extensive mechanistic research has been conducted, the optimal management of IMR remains elusive. Clinical studies in the past often included MR of multiple etiologies, including degenerative or nonischemic origin and IMR grouped into the same category, leading to confusion and incorrect conclusions regarding the natural history and true long-term impact of IMR. It is important to distinguish IMR from mitral regurgitation resulting from nonischemic etiologies. Mitral regurgitation is often associated with coronary artery disease without a direct cause-and-effect relationship. Given the prevalence of coronary artery disease, the association of myocardial infarction and nonischemic mitral regurgitation is a common clinical association. IMR must be distinguished from mitral insufficiency caused by degenerative, rheumatic, congenital, and infectious etiologies, as well as that arising from idiopathic dilated cardiomyopathy. Our recognition of IMR as a distinct clinical entity has improved the understanding of the pathophysiology and natural history of this distinct disease process.

IMR can produce a variety of very variable clinical presentations depending on infarct characteristics and subsequent postinfarction ventricular remodeling. The size, location, and transmurality of the myocardial infarction (MI) sets in motion varying degrees of left ventricular remodeling that subsequently determine the severity, time course, and clinical manifestations of IMR. The presentation may be either acute and immediately life threatening, or develop insidiously over time in association with congestive heart failure (CHF).

The advancement in catheter-based intervention for coronary artery disease and medical therapy has resulted in dramatically improved survival from ischemic heart disease and acute myocardial infarctions (AMI). Although survival following an AMI has improved, there has been a subsequent increase in the incidence of congestive heart failure. The American Heart Association estimates that 5.7 million Americans were in various stages of congestive heart failure in 2006.1 An increasing number of ischemic cardiomyopathy patients will develop left ventricular dysfunction and IMR as a result of annular dilatation and/or papillary muscle displacement.

Early after an acute myocardial infarction (AMI), between 11 and 55% of patients develop clinical (mitral systolic murmur), echocardiographic, or angiographic evidence of IMR. A recent echocardiographic study investigating the incidence of mechanical complications after acute myocardial infarction found a significantly reduced incidence of mitral regurgitation in the percutaneous coronary intervention (PCI) era, but still noted a very large incidence of post-MI acute ischemic MR, 28 versus 53%.2 Many of the murmurs early after acute myocardial infarction are transient and disappear by the time of discharge without further intervention, implying a correction of acute MR without intervention.

In one study, 19% of 11,748 patients who had elective cardiac catheterization for symptomatic coronary artery disease (CAD) had ventriculographic evidence of mitral regurgitation (MR).3 In most of these patients the degree of mitral insufficiency was mild, but in 7.2% the degree of regurgitation was 2+ or greater, and in 3.4% MR was severe with evidence of heart failure.3 In another study of consecutive cardiac catheterizations, 10.9% of 1739 patients with CAD had MR.4

Collectively, these data indicate that IMR is frequent early after AMI, but in many patients it is mild or disappears completely. The relatively high incidence of IMR in catheterized patients with symptomatic CAD suggests that chronic IMR persists in many patients after acute infarction and/or may subsequently develop in others. Approximately 16.8 million Americans have CAD, 7.9 million have had at least one myocardial infarction, 9.8 million suffer from angina, and millions more have asymptomatic coronary atherosclerosis that remains to be diagnosed.1 Using these figures, the incidence of IMR in the United States is estimated to be 1.2 to 2.1 million patients, with approximately 425,000 patients having moderate or severe IMR with heart failure.3,5

IMR has traditionally been classified based on the time course of clinical presentation. Acute IMR occurs in the immediate postinfarction period, and the patients are often in hemodynamic distress. Fortunately, this population of patients represents a minority of the IMR cohort. Chronic IMR represents the clinical presentation for the majority of patients with IMR. The chronic presentation is one of progressive and insidious nature and is often associated with decremental decrease in left ventricular function.

In his landmark American Association for Thoracic Surgery–invited distiguinshed lecture and corresponding 1983 paper, Cardiac Valve Surgery: The French Correction, Dr. Carpentier provided an insightful way to mechanistically approach MR, including IMR. This classification scheme categorizes the improper leaflet coaptation of mitral regurgitation into three subsets based on leaflet and chordal pathology. In type I dysfuction, leaflet motion is normal and mitral regurgitation results from mitral annular dilatation. Type II is mitral regurgitation as a result of leaflet prolapse or excessive motion. In type III dysfunction, there is leaflet restriction or tethering, and is further subclassified into "a" (tethering during diastole) and "b" (tethering during systole and diastole).

IMR can result from type I, II, or IIIb dysfunction. Acute postinfarct mitral regurgitation can be a result of type II dysfunction, in which there is papillary muscle rupture. However, acute IMR is more often associated with much more subtle changes in the mitral valve apparatus. Chronic IMR classically presents as either type I or IIIb dysfunction. Pure annular dilatation with normal leaflets (type I) results from left ventricular remodeling and its geometric sequalae on the mitral valve apparatus. Type IIIb is the more common dysfunction associated with chronic IMR, and is the result of papillary muscle displacement and subsequent tethering of leaflets that results from ventricular dilatation and remodeling that occurs with ischemic cardiomyopathy. Often, a combination of type I and IIIb dysfunctions are seen in patients with chronic IMR and cardiomyopathy.

Normal Valve Function

The mitral valve has six anatomical components: leaflets, chordae tendineae, annulus, papillary muscles, left ventricle (LV), and left atrium. The mitral annulus is saddle-shaped (hyperbolic parabloid with two-directional curvature) with cephalad promontories near the mid portions of the anterior and posterior leaflets and caudad depressions at the commissures (Fig. 29-1).6 This unique shape, which is present in all mammalian mitral valves, has been shown, using finite element analysis, to reduce leaflet, annular, and chordal stress.7 Function of the normal mitral valve is complex and involves precisely timed interactions among the six components. These interactions are most easily described by relating the changes in each of the six components during a cardiac cycle. For this description the cardiac cycle is divided into four periods: systole, diastole, isovolumetric relaxation (IVR), and isovolumetric contraction (IVC), as defined in the following.

End systole (ES) is defined as the maximum negative left ventricular dP/dt and end diastole (ED) is defined as the peak of the QRS complex. End isovolumetric contraction (EIVC) is defined as the first time point at which the aortic root dP/dt is greater than zero. End isovolumetric relaxation (EIVR) is defined as the time at which the LVP is 10% of LVPmax and the left ventricular dP/dt less than 0.

During isovolumetric contraction (IVC), left atrial filling begins immediately after the mitral valve closes and before the aortic valve opens. Flow through the mitral valve briefly reverses as the leaflets coapt and bulge toward the atrium.

During systole the left atrium rapidly fills and reaches maximum volume near ES. The position of the annulus ascends (away from the apex) slightly during atrial contraction, which occurs during late diastole, does not change during IVC, and descends progressively 1 to 1.5 cm toward the apex throughout systole. The annulus asymmetrically contracts during atrial and ventricular systole and in humans reaches a minimal area (mean reduction of 27%) in mid systole.8–10 Immediately after atrial contraction, the mitral leaflets approach each other and close within 20 to 60 milliseconds after pressure crossover when LV pressure exceeds LA pressure. Because the total area of leaflet tissue is approximately twice the total area of the annulus, the apposition point of the two leaflets at the time of pressure crossover is very near the plane of the annulus. At closure, approximately 30% of the anterior cusp and 50% of the longer posterior cusp are in apposition. Chordae tendineae attached to the free edges and body of the leaflets restrict the upward movement of the slightly compliant leaflets and produce a tight seal along the line of apposition. Chordal tension peaks in early systole and begins to fall slowly in late systole and rapidly during IVR. Papillary muscles begin to shorten during late IVC and throughout systole in synchrony with shortening of the adjacent ventricular wall. The actual distance the papillary muscles shorten is small and ranges between 2 and 4 mm.11,12 During systole, the directions and timing of left ventricular contraction are not necessarily uniform throughout because of the complex anatomical arrangement of muscle bundles and the timing produced by the impulse conduction pattern.13 LV shortening is greater in both equatorial axes than in the long axis.13 LV wall thickness increases during IVC and decreases rapidly during IVR.14 Peak wall thickness occurs near ES, but timing of the exact peak varies slightly between ventricular wall segments. During systole the ventricle progressively twists counterclockwise (as viewed from the apex) along its longitudinal axis to reach a maximum at ES.15

During IVR the left atrium begins to empty when left atrial pressure crosses over and exceeds LV pressure. The atrium empties rapidly in early diastole and further diminishes with atrial contraction in late diastole just before IVC. The LV may actually generate negative pressure in early diastole if left atrial pressures are low. During IVR, the mitral annulus to LV apex distance lengthens as the mitral annulus ascends in early diastole, descends slightly, and ascends again with atrial contraction. The area of the mitral orifice increases slightly during IVR and continues to increase during diastole until it reaches a maximum just before the left atrium contracts. In humans, the annular area index reaches a maximum of 3.9 ± 0.7 cm2/m2.10 As the annular area increases, its shape changes asymmetrically; most of the area increase is caused by lengthening in the posterior and lateral parts of the annulus (away from the fibrous trigone).8 During IVR the mitral leaflets separate approximately 30 milliseconds before left atrial pressure exceeds LV pressure.16 Peak blood flow through the valve occurs early in diastole, but the mitral leaflets reach their maximal open position before peak flow occurs and begin closing while flow is still accelerating.

The papillary muscles may shorten very slightly during early IVR, but do not begin to lengthen until the beginning of diastole. Papillary muscles reach maximum length shortly after ED during IVC. Chordal tension decreases rapidly during IVR and remains near zero until late diastole, when a small increase occurs. The left ventricle relaxes and dilates after ES and reverses the complex deformations of LV shape produced by systole. During early diastole, and the period of rapid filling, the ventricle dilates primarily along both equatorial axes and much less along the longitudinal, base-to-apex axis. Only a little shape change occurs in mid diastole and after atrial contraction. Ventricular wall thickness decreases primarily during IVR. Last, the ventricle rapidly untwists (rotating clockwise) during early diastole and more gradually during mid and late diastole.

Mechanism of Ischemic Mitral Regurgitation

Acute Ischemic Mitral Regurgitation

Papillary muscle rupture is a rare, often fatal complication, occuring in 1 to 5% of patients who die after myocardial infarction, as determined by autopsy studies.17 The rupture involves the posteromedial papillary muscle in approximately two-thirds of acute, severe IMR. The vascular supply to the posteromedial papillary muscle is dependent on a single coronary artery (the right coronary artery, or the circumflex coronary artery in a left-dominant system), whereas the anterolateral papillary muscle receives dual supply from the left anterior descending and circumflex coronary arteries. Papillary muscle rupture results in prolapse of the mitral leaflets, often causing severe acute IMR, hemodynamic instability, and acute pulmonary edema. Acute mitral regurgitation in this setting carries a poor prognosis with a high mortality rate.

More often, acute IMR occurs with subtle changes in the mitral valve apparatus in the absence of leaflet prolapse. Annular dilatation has traditionally been regarded as the mechanism of acute IMR. Laboratory work involving an ovine model from the Stanford group has demonstrated the importance of the septolateral (SL) dimension of the mitral annulus in the pathophysiology of IMR.18–20 The SL dimension, the vertical distance of the mitral annulus from the middle of the anterior annulus to the middle of the posterior annulus, is clinically referred as the anteroposterior (AP) dimension. Timek and coworkers demonstrated that correcting SL distance alone may abolish acute IMR.19

However, other experimental studies have demonstrated subvalvular geometric changes as important contributors to the pathophysiology of acute IMR. With ischemia or infarction, the posterior papillary muscle elongates 2 to 4 mm in the sheep and dog12,21; the tip moves 1.5 to 3.0 mm closer to the annulus.12,22 In the sheep model of acute IMR, the uninfarcted, preserved anterior papillary muscle contracts earlier and more vigorously than before infarction. This moves the tip 4 to 5 mm further away from the annular plane at mid systole than before infarction.23 The Stanford group has meticulously characterized this discoordination of normal synchronous papillary muscle contraction and its complex effect on leaflet coaptation.24,25 Tibayan and coworkers demonstrated that by using mitral suture annuloplasty (Paneth-type), altered annular and subvalvular geometry can be corrected, and acute IMR is abolished.26 Neilsen and coworkers in a pig model of acute IMR demonstrated that imbalanced chordal force distribution is directly related to the development of acute IMR.27 The authors showed a decrease in tension of the primary chorda from the ischemic posterior left ventricular wall to the anterior leaflet, whereas the tension of the chorda from the nonischemic anterior left ventricluar wall to the anterior leaflet increased.

These data suggest that acute IMR is a result of complex interaction of small changes in the mitral valvular complex, and not merely simple annular dilatation, as previously believed.23,28,29 These small changes are difficult to demonstrate by standard imaging techniques, making repair challenging in this acute setting. In addition, finite element analysis of the mitral annulus has suggested an association between the normal annular saddle shape of the mitral valve and its competency.30

Chronic Ischemic Mitral Regurgitation

In chronic IMR, mitral valve prolapse has been described in occasional patients, but the vast majority have incomplete mitral valve closure owing to left ventricular remodeling and papillary muscle displacement. Remodeling often manifests as annular dilatation with papillary muscle and chordal restriction of leaflet motion. Pathologic studies consistently show fibrosis and atrophy of infarcted papillary muscles, and none demonstrate papillary muscle or chordal elongation.31–34 Nevertheless, surgeons describe elongated chordae in some patients with myocardial infarction and mitral regurgitation. Elongated chordae and mitral valve prolapse without mitral regurgitation probably antedate the infarction in these patients.35 In a patient with preexisting mitral valve prolapse, ventricular infarction may cause the previously competent valve to leak. This hypothesis explains sporadic observations of mitral valve prolapse in patients with chronic IMR, but requires preinfarction echocardiograms for confirmation.

Ovine experiments using sonomicrometry array localization to study postinfarction MR that evolves during the first 8 weeks after infarction have added insight relevant to the pathogenesis of chronic IMR. As a result of ischemic left ventricular remodeling, a combination of asymmetric annular dilatation and leaflet tethering by both papillary muscles occurs to alter the normal saddle shape of the annulus and produce chronic IMR.36 Previously thought to be fixed, the anterior portion of the annulus does dilate.37 The intertrigonal distance has been shown to increase, in addition to the increase in the SL dimension.38 The annular area dilates by at least 60% at all time points during systolic ejection, but the dilatation involves all of the muscular annulus. The posterior or mural portion of the annulus directly adjacent to the infarct moves away from the relatively fixed anterior commissure (at the anterior fibrous trigone) and stretches the anterior portion of the mural annulus and the posterior portion of the aortic-based annulus, which are remote from the infarct. This finding illustrates how a moderately sized (21% of the LV mass) localized infarct remodels and distorts remote, uninfarcted myocardium including the mitral valve annulus (Fig. 29-2A).39

Lateral displacement of the posterior papillary muscle appears to play a major role in the development of chronic IMR.38 Interestingly, the posterior papillary muscle tip to posterior commissure relationship does not change significantly. Both these points are displaced together, away from the relatively fixed anterior commissure as a result of the remodeling process (Fig. 29-2B). This indicates that the posterior papillary muscle tethering is more pronounced at its most anterior connection with both leaflets near the center of the coaptation line and not at the commissure. The anterior papillary muscle tip is displaced significantly from both commissures but further from the posterior commissure. This indicates that the tethering effect of the anterior papillary muscle is greatest along both leaflets from the anterior commissure to the middle of the coaptation line. Together, these findings suggest that in this model, the postinfarction ventricular remodeling process tethers the anterior portion of both leaflets.

The concept of leaflet tethering as a contributing factor in the pathogenesis of chronic IMR is not new. Two echocardiographic reports, one studying the same sheep model presented here and one human study, demonstrated findings consistent with the experimental results cited in the preceding. Otsuji and coworkers applied a very effective three-dimensional echocardiographic technique to quantify leaflet tethering in the same ovine model.40 These authors also reported mid-systolic distortions between both papillary muscle tips and the anterior commissure, but did not observe changes between papillary muscle tips and the posterior commissure or annular dilatation. Yiu and coworkers used quantitative two-dimensional echocardiography to corroborate these findings clinically by comparing normal controls with a cohort of patients with varying degrees of chronic IMR.41 They found that ventricular distortions, which most closely correlated with the degree of mitral regurgitation, occurred between the posterior papillary muscle tip and anterior commissure and posterior displacement of the anterior papillary muscle tip.

To summarize, the geometric changes that lead to acute IMR are multiple but extremely subtle (<5 mm) and are not reliably imaged by currently available clinical modalities. Chronic IMR involves larger changes (1 to 2 cm) that cause moderate annular dilatation and complex leaflet tethering along the anterior and mid leaflet coaptation line. It is a process resulting from complex geometrical alteration of the mitral valve apparatus as a result of ischemic left ventricular remodeling (Fig. 29-3).

Acute Ischemic Mitral Regurgitation

Clinical Presentation

Clinically, acute IMR usually presents abruptly with chest pain and/or shortness of breath. The presentation is that of an acute myocardial infarction and occasionally may be silent. These patients are often hemodynamically unstable, in cardiogenic shock. They often present in extremis with symptoms of congestive heart failure, including pulmonary edema, systemic hypotension, oliguria, acidosis, and poor periperhal perfusion. On physical examination, in the setting of profound MR, most patients have a loud apical, holosystolic murmur that radiates to the left axilla; but with lesser degrees of MR, a murmur is inconsistently detected.

Nearly all electrocardiograms are abnormal,17,42,43 but only slightly more than half are diagnostic of AMI. Some of the nondiagnostic changes include right or left bundle branch block and nonspecific ST- and T-wave changes in the anteroseptal, lateral, or inferior leads.17,42,44 Most patients are in sinus rhythm. In autopsy series, the incidence of subendocardial infarctions is approximately equal to the incidence of transmural infarctions.17,43 Frequently patients with papillary muscle rupture have electrocardiographic evidence of an inferior MI. When the ECG is diagnostic, inferior wall infarctions are much more common than anterior and lateral wall infarctions. Conduction abnormalities are relatively uncommon and are more often found in patients with postinfarction ventricular septal defects. Chest x-rays nearly always show signs of pulmonary congestion, interstitial pulmonary edema, and pulmonary venous engorgement.42 Cardiomegaly is usually not present, and there is usually no sign of left atrial enlargement in acute MR.17

The differential diagnosis for acute IMR includes postinfarction ventricular septal defect, massive AMI without significant MR, and ruptured chordae tendineae without AMI. Right heart catheterization usually shows elevated pulmonary arterial pressures with prominent V waves reaching 40 mm Hg or higher.42,45 Mean pulmonary artery wedge pressures are greater than 20 mm Hg unless cardiac output is very low. Mixed venous oxygen saturations are often well below 50% and reflect low cardiac output with indices that range from 1.0 to 2.9 L/m2/min.45 In the presence of a loud systolic murmur, absence of an oxygen step-up in the pulmonary artery is strong evidence against the diagnosis of postinfarction ventricular septal defect. Electrocardiographic evidence of AMI distinguishes acute IMR from acute chordal rupture, but the lack of EKG changes does not differentiate the two diseases.42

Transthoracic echocardiography (TTE) assesses the degree of MR, confirms wall motion abnormalities, and may demonstrate flail mitral leaflets. However, transesophageal (TEE) echocardiography is the diagnostic imaging tool of choice. This modality definitively documents the degree and characteristics of the MR jet of IMR, associated wall motion abnormalities, and the status of the posterior papillary muscle. In the modern era, the information provided by TEE is vital in making an accurate decision to repair versus replace the mitral valve. Typically the left atrium is not enlarged, but the left ventricle shows signs of volume overload and segmental wall motion abnormalities. Color flow Doppler velocity mapping documents the presence of MR after myocardial infarction and semiquantitates its severity. Ejection fractions vary widely, but do not reflect the extent of the LV infarction.

Despite hemodynamic instability, most patients have a diagnostic cardiac catheterization primarily for definition of coronary arterial anatomy. However, the wisdom of prescribing cardiac catheterization for patients in cardiogenic shock is highly questionable in that revascularization of obstructed, remote coronary vessels is not likely to improve a patient's chances for immediate survival. Approximately half of catheterized patients have single-vessel disease, most often of the right coronary artery.42,45 Most of the remainder have three-vessel disease.17,42–44 Ventriculography shows increased LV volume at both end diastole and end systole, severe MR, segmental wall motion abnormalities, and a wide range of ejection fractions, which are generally over 40% and frequently over 60%.44,45 LV end-diastolic pressures are usually elevated, with prominent left atrial V waves and moderate pulmonary hypertension. Occasionally patients may have mild or moderate tricuspid regurgitation. Cardiac output is usually very low in this disease process.

Indication for Surgery

Immediate surgery is the best chance for survival for most patients with acute, severe postinfarction MR. The goal of medical therapy is to stabilize the patient and optimize hemodynamcs in preparation for surgical intervention. A few, highly selected patients without papillary muscle rupture early in their presentation have been treated by emergency PCI and/or thrombolytic therapy in an attempt to reduce the size of the infarct and thereby reduce mitral regurgitation.46–49 PCI or thrombolysis carried out within 4 hours of the onset of AMI may on occasion produce spectacular reversal of both the infarction and mitral regurgitation.47–49 However, less rapid PCI may not succeed in preempting the infarct and aborting mitral regurgitation.49 PCI and thrombolysis in catheterized patients are potentially worth trying if patients reach medical attention soon after the onset of symptoms, are sufficiently stable, and can be followed by echocardiography. However, in many patients PCI and thrombolysis do not provide a favorable outcome.46 In one study, 17% of patients with acute IMR and successful thrombolysis died in the hospital; in those with successful PCI, 50% died shortly afterward, and 77% were dead in 1 year.46 Of the survivors, the majority continue to have 3+ or 4+ MR.46

For patients who have acute postinfarction angina with 1+ or 2+ MR, urgent myocardial revascularization is indicated to relieve angina and prevent extension of the infarction. It is important to prevent progression of MR and the development of congestive heart failure or cardiogenic shock. This is usually accomplished by thrombolysis or PCI. If these measures are unsuccessful, operation is rarely completed in time to reverse the infarction, but early operation may reduce the size of the ultimate infarct.50–52 The presence of mild to moderate IMR does not increase operative mortality, but the presence of congestive heart failure is a risk factor.53 In these patients, the mitral valve is generally not addressed unless intraoperative transesophageal echocardiography indicates 3+ or 4+ MR.

Indications for emergency surgery for acute, severe postinfarction MR vary among institutions and probably explain wide discrepancies in reports of hospital mortality.54,55 In this group of patients, medical therapy does not produce survivors and patients denied operation are not reported.42,44 Aged patients are less likely to survive operation, and there are only anecdotal reports of successful operation in octogenarians.47,54,56 Other risk factors for hospital death are severe congestive heart failure, the number and severity of comorbid diseases such as renal or pulmonary problems, presence of an intra-aortic balloon pump, reduced ejection fraction, and greater number of diseased coronary arteries.57

Surgical intervention for acute, severe postinfarction MR consists of mitral valve repair or replacement with or without myocardial revascularization. Nearly all surgeons recommend revascularization of all significantly obstructed coronary vessels away from the site of the infarction.55 Improved methods of cardioplegia support this recommendation even in patients with preoperative cardiogenic shock who have had cardiac catheterization. The wisdom of blind revascularization of remote coronary vessels in patients who have not had preoperative cardiac catheterization and revascularization of the infarct artery more than 4 to 6 hours after onset of pain is less clear.58 On a statistical basis, only half of patients with acute IMR have multivessel coronary artery disease.17,42,45 Revascularization of completed infarctions favorably influences subsequent ventricular remodeling.51,52,59

With echocardiographic findings amenable to repair, a simple, central regurgitant jet, minimally tethered leaflets, and no papillary muscle pathlogy such as rupture, a simple repair with an annuloplasty ring can be enterained. However, replacement of the diseased valve should be considered if the effectiveness and durability of the repair are at all in question. These patients are often critically ill, and will not tolerate the increased cardiopulmonary bypass time associated with mitral valve replacement after failed mitral repair. When performing mitral valve replacement, it is important to preserve the chordal attachments to the annulus to limit adverse ventricular remodeling (Fig. 29-4). Bioprosthestic valves are a reasonable choice since anticoagulation is not benign and durability is a relatively minor concern in these patients with poor long-term prognosis.60,61

Results

Published results of mitral valve replacement for acute IMR are poor. Hospital mortality ranges from 31 to 69% and probably reflects the selection process more than quality of care. Variables that increase early mortality include patient age, cardiogenic shock, comorbid conditions, the amount of infarcted myocardium, and delay in operation.42,44,45,54 More recent experience may be better because of prompt diagnosis, early surgery, complete revascularization, and application of chordal preservation techniques that better preserve left ventricular function.60–67 Several techniques are available for preserving chordae (see Fig. 29-4).62,68,69

Controversy exists over whether the mitral valve should be repaired or replaced in the setting of acute IMR. Repair of the valve in acute IMR potentially enhances the complexity of the operation as well as the cardiopulmonary bypass time as compared with mitral valve replacement. As demonstrated, the anatomical derangements in the acute IMR setting may be very subtle, resulting in annular dilatation and malcoaptation of the leaflets, in which case an undersized ring annuloplasty is a reasonable option. Long-term (5-year) survival in patients who survive the perioperative period is poor and even in modern reports hovers around 50%.60,61 Even with poor prognosis, we believe mitral valve annuloplasty ring is a reasonable option for pathology amenable to repair.

Papillary muscle rupture is less common as the etiology for acute ischemic mitral regurgitation, and this subset of patients is usually much less ill. A recent retrospective analysis of post-MI papillary muscle rupture patients noted a marked improvement in postoperative survival when either valve repair (reimplantation of the papillary muscle and annuloplasty) or replacement was performed in combination with concomitant revascularization. It is encouraging that the authors noted normalization in 5-year survival (79%) and freedom from congestive heart failure to that of matched MI controls without papillary mucle rupture (28 ± 8% versus 36 ± 6%) was noted, predicting an optimistic outcome for this patient population in the modern surgical era.70

Chronic Ischemic Mitral Regurgitation

Clinical Presentation

Chronic IMR represents the majority of patients with IMR. Between 10.9 and 19.0% of patients with symptomatic coronary artery disease who have cardiac catheterization3,4 and 3.5 to 7.0% of patients who have myocardial revascularization have IMR.71–74 The majority of these patients have 1+ to 2+ mitral regurgitation, without evidence of heart failure.3,4,71–73 In patients with chronic IMR, three major variables interrelate to produce the clinical spectrum of patients with varying combinations of symptomatic ischemia and heart failure. As with acute IMR, the three major variables are: (1) the presence and severity of ischemia, (2) the severity of mitral regurgitation, and (3) the magnitude of left ventricular dysfunction. First, patients with obstructive coronary artery disease may be asymptomatic or have stable, progressive, unstable, or postinfarction angina or its equivalent. Because of disabling symptoms, threat to left ventricular mass or statistically shortened survival, ischemia is a compelling problem that must be addressed therapeutically. The approach and methods do not differ from similar patients who do not have IMR.

The second variable is the severity of mitral regurgitation. At present, 1+ or 2+ mitral regurgitation in patients without symptoms of heart failure does not compel invasive therapy for mitral regurgitation. Severe mitral regurgitation and/or symptoms of heart failure require evaluation for possible operation irrespective of the therapy needed for ischemia. The third variable is the degree of left ventricular dysfunction. LV dysfunction is the most difficult to assess in the presence of mitral regurgitation. Symptoms of heart failure may be caused by left ventricular dysfunction secondary to ischemia, mitral regurgitation, or both.

The primary purpose of diagnostic studies is to determine the severity of coronary artery disease and its anatomy, the severity and mechanism of mitral regurgitation, and the degree of left ventricular dysfunction. In chronic IMR, the ventricular geometry and function reflect remodeling owing to both the mitral regurgitation and infarction; therefore, these patients may require diagnostic studies and perhaps operative procedures that are different from patients with CAD associated with mitral regurgitation. It is also important to define comorbidity of other organ systems by appropriate diagnostic studies dictated by the patient's history, physical examination, and screening laboratory findings.

In patients with IMR, the ECG usually shows evidence of a prior myocardial infarction. The incidence of arrhythmias varies but atrial fibrillation as noted earlier is quite common. In patients without failure and mild mitral regurgitation, heart size by chest x-ray is normal or slightly enlarged; the left atrium is seldom enlarged. In those with moderate or severe mitral regurgitation and/or severe left ventricular dysfunction, the heart is enlarged and usually the left atrium is also enlarged.

Transthoracic echocardiography and TEE are useful in determining the etiology of mitral regurgitation. Two-dimensional echocardiography reliably detects ruptured chordae, annular calcification, and myxomatous degeneration, which are not features of chronic IMR, and differentiates rheumatic valve disease, endocarditis, and degenerative valve disease. Echocardiography also effectively assesses regional wall motion abnormalities and global left ventricular function. The degree of mitral regurgitation is also well quantified by color flow Doppler measurements. Characteristics of the mitral regurgitation demonstrated by TEE will also aid in the decision to repair or replace the mitral valve. Extremely dilated annulus with severe bileaflet tethering (Carpentier Type IIIB) and eccentric or complex regurgitation jet are echocardiographic findings that repair may not be effective or durable.

Cardiac catheterization provides information regarding the coronary arterial anatomy and pathology. Ventriculograms add little to the data provided by TTE or TEE and should be avoided, especially in patients with impaired renal function. Measurements of chamber pressures and estimates of cardiac output contribute to the overall evaluation of left ventricular function. Pulmonary hypertension, when present, is typically moderate and correlates with the degree of left ventricular dysfunction and/or severity of mitral regurgitation.

Indications for Surgery

The decision to intervene surgically in IMR can be a challenging process. Chronic IMR represents a set of patients with symptoms of insidious onset and often of chronic duration. Patients may primarily present with symptoms of ischemic coronary artery disease, and on preoperative workup are found to have significant concomitant mitral regurgitation. Although this group of patients may have preserved left ventricular function, compromised ventricular function is not uncommon. The surgical indication in this group of patients is coronary artery disease requiring CABG with coexistent significant mitral regurgitation. Important factors to consider in the decision to intervene on mitral regurgitation include the following: the impact of coronary artery bypass grafting (CABG) alone on the progression of IMR; the impact of CABG with or without MVR on survival; the additional risk of MVR at the time of CABG; and the choice of valve repair and replacement.

Other patients may primarily present with symptoms of congestive heart failure and signs of significant mitral regurgitation. Preoperative catheterization subsequently may reveal significant CAD. These patients tend to have more compromised left ventricular function compared with the first group. The indication for surgery is congestive heart failure. In reality, most patients present in a clinical spectrum somewhere along the two clinical scenerios. In this section, however, the indications for surgery of each scenerio are discussed separately.

Ischemic Mitral Regurgitation and Coronary Artery Disease

The optimal management of concomitant IMR at the time of CABG remains to be determined. Most surgeons agree that concomitant severe (4+) mitral regurgitation should be addressed at the time of CABG and that revascularization alone will not ameliorate severe mitral regurgitation. Similarly, most surgeons agree that trace to mild (1+) mitral regurgitation should be left alone because it will not adversly affect long-term symptomatology or prognosis. Yet, the optimal management of mild to moderate (2+) mitral regurgitation remains controversial.

Multiple factors must be addressed in deciding on the management of IMR at the time of CABG. First, IMR has been demonstrated to have a negative impact on long-term survival. Numerous studies have documented the negative impact of IMR on long-term survival after an acute MI.75–78 Similarly, in patients undergoing PCI for acute coronary syndromes, IMR has been demonstrated to have a negative impact on survival. Three-year survival in this group ranges from 46 to 76%, based on the severity of mitral regurgitation.79 Moreover, in the setting of CABG, myocardial revascularization alone in patients with chronic IMR has a higher hospital mortality than in patients without IMR.74 Mild (1+) IMR increases operative mortality from 3.4 to 4.5%72–74,80 and moderate (2+) IMR raises operative mortality from 6 to 11%.72–74,81 Two-year survival for revascularization alone in patients with 1+ and 2+ mitral regurgitation is 78 and 88%, respectively.82 Five-year survival rates for patients with mild mitral regurgitation range between 70 and 80%.3,72,83,84 For moderate mitral regurgitation, 5-year survival ranges between 60 and 70%.85,86 Many surgeons argue that these data suggest that concomitant IMR should be addressed during CABG to affect survival.

Those who advocate the conservative approach of revascularization alone, without treatment of the IMR, argue that revascularization will improve regional wall motion abnormalities, papillary muscle function, and potentially correct IMR.71,87,88 Moreover, there are data that suggest that survival and long-term functional status are not improved with concomitant MVR,89,90 thus bringing into question the benefit of the higher operative risk associated with simultaneous MVR. Surgeons who advocate mitral valve repair/replacement for moderate IMR during CABG cite studies that suggest revascularization does not correct IMR,91 and that uncorrected IMR may result in late symptoms and decreased long-term survival.76,82 Furthermore, mitral valve repair with annuloplasty ring is nearly always technically feasible, obviating the need for valve replacement and the associated anticoagulation of mechanical valves or subsequent valve replacement required of bioprosthetic valves. Operative risk with combined CABG/MV repair today is much better than the older series, with operative mortality in the 3 to 4% range.92–94 Kron and collegues have concluded that the addition of MVR does not increase the operative risk of CABG in ischemic cardiomyopathy.95 Considering the increased complexity associated with redo sternotomy with patent grafts for MVR for recurrent or progressive IMR, some feel aggressive management of IMR is justified on initial sternotomy and revascularization, given simplicity of repair when compared with subsequent redo-operation.

The impact of isolated CABG without MVR on the progression of moderate IMR has been examined. Previous studies suggest that CABG alone improves IMR grade and functional status.71,87,88 However, in contrast recent reports have suggested that CABG alone is not the optimal therapy for moderate IMR.91,96,97 A study from the Cleveland Clinic Foundation reported that moderate (2+) IMR does not resolve with CABG alone, and furthermore, is associated with reduced survival.98 Between 1980 to 2000, 467 patients with moderate IMR underwent CABG alone. Longitudinal analysis of 267 follow-up echocardiograms from 156 patients demonstrated that early postoperative improvements in IMR did not persist. IMR of moderate or greater severity was present in 60% of patients by postoperative week 6. Interestingly, postoperative IMR severity was not predicted by cardiac function or extent of CAD. Furthermore, early survival of patients with unrepaired moderate IMR was reduced compared with similar patients without IMR. Based on their results, the authors concluded that a mitral valve procedure is warranted for such patients at the time of CABG. A retrospective registry review of the progression of mitral regurgitation following isolated coronary artery bypass surgery, examined 438 patients with preoperative mitral regurgitation of less than or equal to 2+ requiring CABG.99 New 3+ to 4+ mitral regurgitation developed in 10% of patients with no prior mitral regurgitation, 12% with pre-CABG 1+ mitral regurgitation, and 25% with pre-CABG 2+ mitral regurgitation, thereby implying significant progression of IMR without treatment. Preoperative left ventricular dysfunction and large left ventricular size were identified as predictors of mitral regurgitation progression post-CABG. Although no correlation was identified between the extent of CAD or the number of grafts performed, incomplete revascularization in the PDA territory was identified as a significant predictor of mitral regurgitation progression, suggesting a role for incomplete revascularization and left ventricular remodeling in the progression of IMR.

To directly address the impact of MVR for moderate IMR, studies have compared the results of CABG alone versus CABG with concomitant MVR in the setting of IMR.100–105 The results do suggest that post-operative mitral regurgitation is improved with CABG with concomitant MVR. However, whether there exists an improvement in long-term survival remains unclear. Several studies suggest no improvement in survival following CABG with concomitant MVR.102,104–107 In contrast, two studies100,101 comparing CABG versus CABG/MVR in patients with CAD and IMR have found a significant improvement in survival. Both studies involved patients with 2+ to 3+ IMR with depressed LVEF. A recent study103 examining 111 patients with moderate to severe IMR and multivessel CAD undergoing either medical therapy, isolated CABG, or CABG/MVR found a greater than 50% reduction in mortality in the CABG and CABG/MVR groups when compared with medical therapy. However, only CABG/MVR independently predicted survival. Moreover, a history of CHF was an independent predictor of cardiac death. Five-year survival has been found to be influenced by the severity of left ventricular dysfunction at the time of operation, age, and comorbid disease.57 Collectively, these studies suggested that concomitant MVR with CABG may improve late survival, especially in the setting of LV dysfunction and CHF, but the definitive answer remains elusive. To clarify the appropriate management of moderate IMR during CABG, a multi-institutional randomized, prospective NIH NHLBI–sponsored clinical trial has begun. In this study, patients with moderate IMR at the time of CABG are randomized to either repair of the MR or isolated CABG alone.

The final factor to address when dealing with IMR is the decision to repair or replace the valve. Gillinov demonstrated the short-term efficacy of mitral valve repair in 97% of patients undergoing elective surgery for 3+ to 4+ chronic IMR.60 Ring annuloplasty was employed in 98% of these repairs and was the sole surgical maneuver on the valve in greater than 80%. There was a distinct inclination in this study to undersize the annuloplasty ring; 79% of the rings were 30 mm or less. Iatrogenic mitral stenosis was not seen even in patients who received 26-mm annuloplasty devices. However, there was no benefit to repair versus replacement in high-risk patients, including older age, higher NYHA functional class, significant wall motion abnormalities, and renal dysfunction. In contrast, Calafiore and colleagues have found no difference in either short- or long-term mortality between MV repair and replacement for patients with 2+ to 4+ IMR.92 Similarly, a recent study found no difference in either operative or 6-year mortality following MV repair as compared with replacement. A prospective, randomized multi-institutional NIH NHLBI–sponsored study is presently underway to definitively determine if there is a difference between MV repair and replacement in the setting of severe ischemic mitral regurgitation.

In summary, patients with CAD with concomitant moderate to severe (3+ to 4+) IMR should undergo CABG/MVR. In patients with moderate (2+) IMR, recent studies may suggest that CABG/MVR may be justified, given the lower rate of morbidity and mortality in the modern surgical era, but this remains to be determined. Left ventricular dysfunction and increased left ventricular dimension along with incomplete revascularization may predict a higher rate of progression of IMR, suggesting CABG/MVR should be performed in this particular group of patients. Patients with symptoms of CHF appear to benefit most from CABG/MVR. Mild (1+) IMR should be left alone, unless: (1) preoperative signs and symptoms are suggestive of periods of more severe mitral regurgitation; and (2) intraoperative TEE demonstrate anatomical findings requiring MVR (ie, significant annular dilatation, leaflet tenting). Most patients with IMR benefit from mitral valve repair. In the most complex, high-risk settings, replacement may be preferable, with no demonstrable difference in survival between repair and replacement.

Ischemic Mitral Regurgitation and Congestive Heart Failure

The ischemic cardiomyopathy patient subset is a complex population to manage. In the current era of aggressive medical management and revascularization, there has been a dramatic decrease in the incidence of acute myocardial infarction and death. However, there is a significant increase in the number of patients suffering from ischemic cardiomyopathy. This disease is often coincident with ischemic mitral regurgitation. As described in previous sections, postinfarction LV remodeling is characterized by progressive LV dilatation with a change to a spherical shape with resultant functional MR secondary to annular dilatation, papillary muscle displacement, and chordal tethering. With functional MR there is increased ventricular volume overload (preload), myocardial wall tension, and LV workload; all of which further contribute to a deleterious cycle of progressive heart failure.

In the past, operative repair of mitral regurgitation was associated with a prohibitively high perioperative mortality, making surgeons cautious about MVR and MVR/CABG in the setting of left ventricular dysfunction. The traditional teaching incorporated the "popoff" valve hypothesis, which believed that the mitral valve functioned as a low-pressure runoff to decompress the failing ventricle. It was believed that surgical correction of mitral regurgitation in the failing ventricle would further contribute to ventricular overload and further contribute to myocardial failure.

However, this hypothesis has been challenged and disproved by Bolling who believes that there is an "annular solution for the ventricular problem."108–111 It is believed that correction of mitral regurgitation alleviates excessive ventricular workload, improves ventricular geometry, and enhances ventricular function; a process referred to as reverse remodeling. The Stanford group has reported in an ovine model of ischemic mitral regurgitation, that reduction of the annulus by an undersized ring reduces the radius of curvature of the left ventricle at the base, equiatorial, and apical levels.112 This decrease in the radius of curvature supports the concept that a small ring can restore a more elliptical shape. Reduction of left ventricular dimensions to a more elliptical shape, reverse remodeling, after annuloplasty have also been demonstrated by clinical studies.113–115

The previously published high operative mortality associated with mitral valve replacement was most likely the result of loss of the subvalvular apparatus, thus indicating the importance of maintaining the integrity of the annular and subvalvular continuity during mitral valve surgery. More recent series have indeed demonstrated much lower perioperative morbidity and mortality (1.6 to 5%), demonstrating safety in carefully selected patients.60,61,116,117

Data from most recently published series have demonstrated improvements in left ventricular ejection fraction, mitral regurgitation, clinical symptoms (NYHA classification), and left ventricular reverse remodeling with surgical intervention.89,100,107–109,118–121 However, questions remain regarding improvements in long-term survival and its overall effect on the natural history of IMR. Some have recently raised questions, despite previous data, about the potential improvement in long-term survival in patients with IMR in the setting of depressed LVEF. In perhaps the largest experience of MVR in patients with severe left ventricular dysfunction, Wu and coworkers examined the impact of mitral valve annuloplasty on mortality risk in patients with mitral regurgitation and left ventricular dysfunction.122 Reviewing their echocardiographic database (n = 682), the authors identified 419 patients who met the criteria for surgical intervention. One hundred twenty-six patients underwent MVR, whereas 263 continued conservative medical therapy. It should be noted that the etiology of mitral regurgitation included both ischemic and nonischemic subtypes. Their analysis demonstrated no improvements in long-term survival in the MVR group versus medical therapy. Talkwalkar and coworkers89 reported a series of 338 patients who underwent MVR. Compared with the control group, patients with depressed left ventricular ejection fraction were more likely to be associated with IMR, concomitant CABG, and NYHA class IV symptoms. Five-year survival was 54%, and when associated with prior CABG, prior myocardial infarction, or concomitant CABG, it was 0, 37, and 63%, respectively.

In agreement with previous reports, these recent studies confirmed the poor prognosis associated with IMR in the presence of LV dysfunction. In Wu's series,122 5-year survival in the setting of IMR and left ventricular dysfunction is less than 50%, regardless of surgical versus medical therapy. A randomized study comparing surgical versus medical therapy for IMR is needed to clarify the benefit of mitral valve surgery in this population of patients.

Patients who present primarily with symptoms of congestive heart failure and IMR provide surgeons with a fairly straightforward indication for intervention. Although it remains unclear if there is a survival benefit with repair, it is clear that there are improvements in symptoms, exercise tolerance, and ventricular remodeling. Given the low operative mortality associated with modern surgical techniques it is reasonable to correct severe IMR in the setting of heart failure. The operative indications for this group of patients are similar to mitral regurgitation of nonischemic origin. Symptoms of CHF and depressed left ventricular function are indications for surgical mitral repair or replacement. The current ACC/AHA valve disease guidelines include a cautious recommendation for consideration of MV surgery in patients with advanced heart failure, but only if MV repair or replacement with chordal sparing are options. Most surgeons agree that concomitant CABG is indicated in the presence of significant CAD.

Results

An undersized annuloplasty ring is effective in the acute correction or reduction of IMR in the majority of patients. However, recurrent mitral regurgitation after annuloplasty has been reported to be present in 30 to 40% of patients. Although these studies are criticized for not addressing the appropriate annuloplasty ring selection and adequately downsizing the ring, both of which have been demonstrated to be vital in adequate long-term correction of IMR. These studies have often used flexible bands or partial rings, which fail to address the dilatation that occurs along the anterior annulus and fail to adequately fix the septal-lateral dimension of the valve.

The Cleveland Clinic reported its series of 585 patients who underwent annuloplasty alone for IMR between 1985 and 2002 in which 678 postoperative echocardiograms were evaulated in 422 patients. The majority of recurrent IMR occurred within the first 6 months. Overall, 28% of patients 6 months after surgery demonstrated 3+ to 4+ mitral regurgitation.123 These results stand in contrast to the findings by Spoor and Bolling, in which minimal recurrent MR was seen up to 4 years after mitral repair with rigid annuloplasty using a ring that was downsized by two sizes.121 It has been found that there is an almost fourfold increase in recurrence rate of IMR with the use of a flexible ring as compared with a nonflexible ring in patients with a preoperative ejection fraction less than 30%.124

Evidence of left ventricular reverse remodeling after annuloplasty ring has been demonstrated in small clinical studies, with excellent results and freedom from recurrent IMR for up to 2 years.113–115 Braun and coworkers reported a series of 87 patients with IMR and depressed left ventricular function (mean LVEF 32%) who underwent undersized mitral annuloplasty and coronary revascularization. Mitral regurgitation grade decreased significantly from a mean of 3.1 to 0.6 at 18 months. Both left ventricular end-systolic and end-diastolic dimensions decreased significantly when compared with preoperative echocardiograms. Interestingly, the left ventricular end-diastolic dimension was found to be the best predictor of reverse remodeling. Left ventricular end-diastolic dimension exceeding 65 mm was associated with a lower probability of reverse remodeling, suggesting a relationship between recurrent IMR and ventricular geometric dimension. Further studies would be useful to identify preoperative echocardiographic parameters predictive of durable mitral valve repair.

Operative results for mitral valve surgery with or without concomitant CABG have improved, with recent series reporting operative mortality of 3-4%.60,61,116,117 However, 5-year survival remains very low, with reported values as low as 30 to 40%, depending on the series (Fig. 29-5).3,72,73,83,125 More recently, Gillinov and coworkers reported 5-year survival in the propensity-matched best risk group of 58% for valve repair and 36% for replacement. This group had significantly fewer NYHA class IV patients and less severe mitral regurgitation preoperatively. In the propensity-matched poorer risk groups (more severe CHF, mitral regurgitation, and emergency surgery) and for the group as a whole, there was no difference between repair and replacement and 5-year survival was uniformly less than 50%.60

There has been no randomized study demonstrating a survival benefit with mitral valve repair/replacement in IMR. In a restrospective analysis of cardiomyopathy including both ischemic and nonischemic etiology, Wu and coworkers reported in their series 5-year survival of less than 50%, regardless of surgical versus medical therapy in patients with mitral regurgitation and left ventricular dysfunction.122 Other series have also demonstrated no survival benefit with mitral valve surgery in patients with IMR.104–106 The similarity of results between surgical and medical therapy points out a need for better understanding of the pathophysiology of IMR and left ventricular remodeling. A randomized study examining the clinical outcome and survival benefit of mitral valve surgery for IMR is warranted.

Given the unstable presentation of patients with acute, severe postinfarction mitral regurgitation, operative intervention is often emergent. A very large percentage of these patients require high-dose pharmacologic support or intra-aortic balloon pump counterpulsation for hemodynamic stablization in the immediate preoperative period.45,58 Although intraoperative monitoring is very institution dependent, typical monitors include electrocardiogram, arterial blood pressure, pulmonary artery catheter with mixed venous oxygen electrode, nasopharyngeal and core body temperatures, and a urinary catheter for urine output. Additionally, a large proportion of centers use intraoperative TEE for optimal management of valvular repair and myocardial optimization.

Chronic IMR is most often performed in an elective manner, except in patients with uncontrolled symptoms of coronary ischemia who may require emergency or urgent operation to prevent myocardial infarction. Preoperative preparation and intraoperative monitors are similar to other cardiac operations. Transesophageal echocardiography and color flow Doppler velocity mapping are considered the standard of care in most centers that perform large volume valve repairs/replacements. After induction of anesthesia, TEE is used to assess: (1) anatomy of the valve; (2) degree of mitral regurgitation, include characteristics of the regurgitant jet; and (3) dimensions and segmental wall motion of the left ventricle. General anesthesia reduces systemic vascular resistance and afterload, thereby masking the degree of mitral regurgitation; therefore, assessment of the valve after administration of vasopressors to enhance afterload may unmask more severe mitral regurgitation. If the amount of mitral regurgitation remains in question or is only mild to moderate, then transfusion after aortic cannulation to increase preload to 1.5 to 2.0 times resting pulmonary capillary wedge pressure may further unmask more severe mitral regurgitation and prompt a decision to expose and repair the valve at the time of other cardiac interventions (ie, CABG). If there is a significant discrepancy between the intraoperative degree of mitral regurgitation and that diagnosed by preoperative TEE, it is probably best to treat according to the preoperative value because this is likely what the patient experiences under normal loading conditions. In patients with marginal left ventricular function, a pulmonary artery catheter with continuous cardiac output and mixed venous oxygen sensing capabilities along with a femoral arterial catheter for rapid intra-aortic balloon insertion, if needed to wean from cardiopulmonary bypass, are recommended.

The most common exposure to the mitral valve is via a median sternotomy, although over the past decade port-access and robotic platforms have afforded many surgeons expertise in exposure to the valve via the right pleural space. If revascularization is required, arterial and venous conduits are harvested before cannulation and initation of cardiopulmonary bypass. Bicaval cannulation is preferred for exposure of the mitral valve. Both antegrade and retrograde cardioplegia are recommended to minimize myocardial stunning and post-bypass ventricular dysfunction. After initiation of cardipulmonary bypass, the aorta is cross-clamped and cardioplegia is given in a standard fashion. If revascularization is needed, the distal coronary anastomoses are performed first to reduce manipulation of the heart and prevent potential atrioventricular disruption after mitral valve repair or replacement.

In patients with a history of CABG, right thoracotomy is an option to avoid injury to previous bypass grafts during a redo sternotomy. Patent grafts, especially the left internal mammary (LIMA) graft to the left anterior descending artery, may be immediately under the sternal table from previous surgery. Because of the inability to isolate and control the LIMA graft from a right thoracotomy, fibrillatory arrest is needed.

Once exposed, the left atrium is opened after developing the interatrial groove, Sondegaard's groove, extending the incision behind the inferior vena cava to the oblique sinus. In cases of redo sternotomy or acute IMR (in which case the left atrium may be small), exposure to the mitral valve can be diffcult and a transeptal approach via the right atrium may be required. Improved exposure may also be obtained by extending the atrial incision superiorly behind the superior caval–right atrial junction. Using a specialized mitral retractor (Cosgrove retractor), the left atrium is lifted and the mitral valve should be exposed for inspection. If further exposure is required the superior vena cava can be transected and the left atriotomy further extended. This is later reanastomosed much like is performed in a bicaval orthotopic heart transplant.

The first and most important step in mitral valve surgery is valve analysis. Inspection will reveal the degree of annular dilatation and leaflet and chordal pathology, and may indicate segments of the mural annulus that appear disproportionately elongated. Traction sutures in the annulus at each commissure elevate the valve and facilitate exposure of the leaflets, chordae, and papillary muscles. Careful inspection should include searching for ruptured, elongated, or sclerosed chordae; fibrotic, atrophied papillary muscle, and redundant or defective leaflet tissue. Most often the entire valve appears normal; sometimes the posterior papillary muscle seems slightly more yellowish brown than the rest of the ventricle and the posterior part of the mural annulus seems slightly elongated.

Echocardiographic evaluation of the valve is important in the decision to repair or replace the valve. A central regurgitant jet with a dilated annulus is suggestive of a successful simple repair with an annuloplasty ring. In contrast, eccentric, complex regurgitant jets with severe bileaflet tethering and papillary muscle pathology such as elongation or rupture suggest that mitral replacement may be the best durable option. In this scenario mitral valve repair may not be a long-term solution. Magne and coworkers have found that preoperative echocardiographic assessment of posterior-leaflet angle greater than 45 degrees to the annulus determines a higher rate of failure of mitral repair using an annuloplasty ring (Fig. 29-6).126

For mitral valves amenable to annuloplasty repair, 2-0 braided suture is routinely used. Often, exposure to posterior and medial portions of the mitral annulus is easiest to achieve. These sutures should be placed first and used to place tension on the mitral annulus to faciliate exposure of the lateral and anterior portions of the annulus. It is essential that the full curve of the needle be used to ensure optimal placement of the sutures in the mitral annulus. Significant tension can develop when the dilated annulus is downsized, so sutures should be placed close together and overlap may be warranted. It is vital that the sutures are placed in the annulus and not in the atrial tissue to avoid subsequent annuloplasty ring dehiscence.

Sizing of the annuloplasty ring can be performed before or after placement of the annuloplasty sutures, depending on the size of the left atrium and adequate exposure. Introduced by Bolling, the concept of reducing annular dilatation with aggressive annuloplasty downsizing aims to optimize leaflet coaptation. When choosing an annuloplasty ring, the intertrigonal distance and/or the surface area of the anterior leaflet are measured. Aggressive downsizing of the annulus (two sizes smaller than predicted by measuring the fibrous intertrigonal annulus) can be achieved with minimal risk of the development of systolic anterior motion, because the posterior leaflet in IMR is often tethered and restricted. Gorman and colleagues have demonstrated that use of a saddle-shaped annuloplasty ring restores normal annular geometry, decreases leaflet strain, and may increase repair durability.127,128 Therefore, it may be prudent to choose a saddle-shaped annuloplasty ring as compared with a flat ring (Fig. 29-7). After the size of the ring is chosen, the annuloplasty sutures are placed in accordance with the geometry of the mitral annulus. Once all sutures are placed, the annuloplasty ring is slid down onto the mitral valve annulus and sutures tied to secure the ring. At this point, a saline test is usually employed to interrogate the competency of the mitral valve repair. If satisfied with the repair, the left atrium is then closed.

If on inspection the valve is not amenable to annuloplasty repair, mitral valve replacement should be performed. Chordal sparing techniques62–64,68,69,129 should be used if the valve is replaced. These methods have significantly reduced postoperative ventricular dysfunction observed after valve excision and have been found to produce no more left ventricular dysfunction than reparative operations.69 Aged patients in sinus rhythm and patients with a life expectancy of less than 10 years who otherwise do not need anticoagulation are candidates for bioprosthetic valves; mechanical valves are recommended for others. The valve is inserted after excising a portion of anterior leaflet63,69 and transposing68 the remaining leaflet tissue to the commissures (see Fig. 29-4). Pledgeted everting sutures (atrial to ventricular, intra-annular placement) are used for intra-annular placement if downsizing and/or low profile valve (ie, mechanical valve) is desired. Noneverting, pledgeted sutures (ventricular to atrial, supra-annular placement) may be used for high-profile bioprosthetic valves or when a friable annulus is encountered. The mural leaflet is plicated into the valve insertion suture line and the struts of the prosthetic valve are placed along the commissural plane to prevent interference with the valve mechanism and left ventricular outflow tract obstruction.

The atriotomy is closed using running sutures, and any proximal coronary anastomoses are completed. Before removing the aortic clamp, all air is evacuated from the ventricle and the mitral valve is kept incompetent using a transvalvular catheter with ventricular and atrial holes. Absence of air and assessment of ventricular wall motion is made by transesophageal echocardiography. Anticipated pharmacologic support is started and satisfactory left ventricular contractility is established before loading the heart. After weaning, cardiopulmonary bypass is restarted. If the left ventricle begins to dilate and wall motion deteriorates, every effort is made to prevent any distention of the ventricle that might reduce myocardial contractile force. This may require placement of a pulmonary artery vent. Decisions for using intra-aortic balloon pump counterpulsation or even temporary left ventricular mechanical assistance are better made early than after multiple attempts to wean from cardiopulmonary bypass have failed.

There have been significant preliminary advances in novel approaches to treating mitral regurgitation, this has included subvalvar approaches, external fixation, and percutaneous repair strategies. Chordal cutting of two critical basal chordae of the anterior leaflet first described by Messas130,131 has been shown to be effective in treating IMR. By relieving leaflet restriction and eliminating the angulation of the anterior leaflet, chordal cutting improves leaflet coaptation while the intact marginal chordae continue to prevent prolapse. Conflicting data from the Stanford group132,133 demonstrate that cutting of second-order chords neither prevented nor decreased the severity of IMR in a sheep. Further investigation is required before this treatment modality is brought into mainstream management of IMR.

Kron and coworkers have reported a successful subvavular repair as an adjunct to standard annuloplasty.134 Direct reposition of the displaced posterior papillary muscle toward the right fibrous trigone appear to ameliorate IMR. In a sheep model, Langer and coworkers confirmed Kron's findings.135 Anchoring a suture at the right fibrous trigone and passing it through the posterior papillary muscle tip and left ventricular wall, the investigators demonstrated reduction of IMR by tightening the suture. Other groups have demonstrated that reapproximation of both papillary muscles may be beneficial in reducing leaflet tethering and IMR.136–138

Recent evidence argues that the development of IMR is a complex, multifactorial process resulting from left ventricular remodeling involving the entire mitral apparatus. Many investigators postulate that adjunctive surgical therapy directed at the remodeled ventricle may prove to be potential therapeutic strategies at treating IMR. Many have stated that IMR is not a valvular process, but a ventricular disease process. In a sheep model, Moaine and coworkers demonstrated that external restraint of the infarct zone attenuated remodeling and reduced chronic IMR.139 Furthermore, external infarct restraint with mesh before infarction prevented left ventricular remodeling compared with the annuloplasty group. The data suggested that left ventricular remodeling associated with infarct expansion may serve as an important therapeutic target. The Acorn cardiac restrain device (CorCap) is a mesh device that is implanted around the heart to reduce wall stress and address left ventricular remodeling. Clinical trials with the Acorn device have shown that it is safe and effective in patients with advanced heart failure and remodeled ventricles, of mostly dilated etiology.140–142 Further investigation is needed before definitive conclusions can be made.

Percutaneous therapies for MR have been attempted in multiple derivations. Rigid coronary sinus–based mitral annuloplasty devices deployed percutaneously have been demonstrated to reduce IMR in animal model.143–145 Subseqent clinical trials have demonstrated clinical feasiblity.146 A percutaneous approach to a modified Alfieri repair utilizing a clip to approximate the anterior and posterior mitral leaflets while creating a double outlet mitral valve has also been clinically tested. Initial results suggest feasibility and an acute reduction in MR.147 Finally, the advent of percutaneous valve replacement technology has expanded to include transcatheter mitral valve prosthesis. Animal studies have been formulated and feasibility remains to be determined. The long-term results of these novel therapies remain to be elucidated. Further work is needed to explore the potential effectiveness and clinical applicability of these new technologies.