It can be reasonably argued that the very dawn of cardiac surgery began with a mitral valve repair. On May 20, 1923, Dr Elliot Carr Cutler (Fig. 35-1) performed the world’s first successful mitral valve repair at the Peter Bent Brigham Hospital in Boston, Massachusetts.¹ Dr Cutler carried out a transventricular mitral valve commissurotomy with a neurosurgical tenotomy knife on a critically ill 12-year-old girl. His choice of instrument was likely influenced by Dr Harvey Cushing who was surgeon-in-chief at the time. A new era in surgery was introduced as well as the reality of mitral valve repair.² Cutler had worked assiduously on this problem in the Surgical Research Laboratories of Harvard Medical School before turning his attention to this critically ill patient. Subsequent attempts at this operation using a device to cut out a segment of the diseased mitral valve resulted in several deaths from massive mitral regurgitation and Cutler eventually abandoned the procedure.³ Of Cutler’s contemporaries, Henry Souttar of England performed a single successful transatrial finger commissurotomy in 1925, but received no further referrals.⁴ After Souttar, there remained little activity in mitral valve repair until the 1940s when Dwight Harken, then the Chief of Cardiothoracic Surgery at the Peter Bent Brigham Hospital, published his groundbreaking series of valvuloplasty patients for mitral stenosis.⁵ Dr Charles Bailey of Philadelphia also published a concomitant series of a similar large group of patients.⁶

That early era focused on mitral stenosis created by rheumatic heart disease, which was extremely common at the time. Surgical treatment of mitral regurgitation for prolapse was first introduced in the 1950s^7-9 but with limited success. Subsequent decades would see the visionary concepts of surgical leaders such as Alain Carpentier, Dwight McGoon, Carlos Duran, and others come to the fore as their visionary ideas for mitral valve reconstruction began to take hold and excellent results were reported. Like many other groundbreaking ideas, their concepts were met with resistance that has gradually dissipated as long-term results by these surgeons have been validated. The concept that repair of mitral regurgitation might serve to further damage a weakened left ventricle (LV) by eliminating the “pop-off” mechanism¹⁶ of the regurgitant valve. This proved a significant barrier to referral that only in the past few decades has been overcome. What has now become firmly established is the significant contribution to overall left ventricular function of the papillary muscle-annular interaction.¹⁷ As a result of these contributions, mitral valve repair, if technically possible, has now become recognized as the procedure of choice for mitral valve pathology of virtually all etiologies, to the extent that mitral valve repair is always considered first in virtually any clinical situation in which the mitral valve is regurgitant.

The bicuspid mitral valve is one of the most complex structures of the human heart, its complexity lies in its multifaceted anatomy. The concept of “form follows function” is particularly applicable to the mitral valve. Because each part of the valve’s anatomy is intimately related to function, there are a variety of pathways whereby regurgitation may be created. If one part of the valvular apparatus fails, regurgitation can result. There are five discrete components to the mitral valve complex: the annulus, the two leaflets (anterior and posterior), the chordae, the papillary muscles, and the LV (Fig. 35-2A).

As part of the fibrous skeleton of the heart, the annulus is the myocardial connective tissue area where the mitral valve leaflets attach to the intersection of the left atrium and LV. It is surrounded by vitally important structures that the cardiac surgeon must avoid for safe surgery: the circumflex coronary artery laterally, the coronary sinus medially, the aortic root superiorly, and the atrioventricular node superior-medially. In myxomatous disease it is the posterior annulus that usually dilates.¹⁸ Previous dictum held that the anterior annulus does not dilate, but recent data suggest that it may dilate a limited amount.¹⁹ Of critical importance to the surgeon are the right and left fibrous trigones. These are intimately related to the anterior annulus and are contiguous with the aortic valve curtain and must be identified during surgery. The trigones, as part of the essential structural framework of the heart, form the anchoring points for ring annuloplasty.

The anterior leaflet of the mitral valve (AML) is in continuity with the left and noncoronary cusps of the aortic valve and is located directly beneath the left ventricular outflow tract (LVOT). It is sometimes referred to as the aortic leaflet. It typically accounts for approximately one-third of the circumference of the annulus, with the posterior leaflet accounting for the rest.²⁰ The posterior mitral valve leaflet (PML) is crescent shaped and is more commonly involved in degenerative disease. For surgical decision making and analysis, both the anterior and posterior leaflets are divided into three parts, corresponding to the three scalloped areas of each leaflet (A1, A2, A3 for the anterior leaflet; P1, P2, and P3 for the posterior leaflet; 1 refers to the leftmost, or lateral scallop; 2 the middle scallop; and 3 the rightmost, or medial scallop; Fig. 35-2B).

There are two papillary muscles, the anterolateral and the posteromedial. Each muscle is attached to the leaflets by the chordae tendinae. These chords are composed of string-like fibrous connective tissue. Primary chords are those that attach to the edge of the leaflet. Secondary chords are those that attach to the underside of the leaflets. Tertiary chords, which are only seen on the posterior mitral leaflet, are those that attach to the undersurface of the leaflet directly from the ventricular wall instead of from the papillary muscle. The papillary muscles each give off chordae to both leaflets and correspond to the anterolateral and posteromedial commissures of the mitral valve. The anterolateral papillary muscle receives blood from both the left anterior descending artery as well as the circumflex artery; the posteromedial one receives blood usually from only the posterior descending artery or a branch of the circumflex artery. Because of its single coronary blood supply, the posteromedial papillary muscle is more susceptible to infarction and rupture than the anterolateral one.

The LV acts in concert with the papillary muscles via the chordae to pull in the leaflet edges during systole, thereby maintaining the line of coaptation and therefore valve competency. If the LV dilates from any etiology, failure of central leaflet coaptation may occur and regurgitation created. Such regurgitation in a valve with normal leaflets but dilated annulus is termed functional mitral regurgitation (MR).

Etiology and Pathophysiology

The underlying etiology of myxomatous disease is a defect in the fibroelastic connective tissue of the valvular leaflets, chordae, and annulus.²¹ The myxomatous defect leads to an abnormal elongation and redundancy of valve tissue and chordae. Each particular anatomic redundancy creates mitral regurgitation in its own particular manner. Annular dilatation (Fig. 35-3) obliterates the normal coaptation line between the anterior and posterior leaflet, causing regurgitation. If primarily posterior annular dilatation occurs, there is a separation in the middle of the valve between the two leaflets and blood leaks through during ventricular contraction. Leaflet redundancy results in a movement of the redundant leaflet into the left atrium during diastole. If severe enough, that movement leads to a compromised coaptation line and MR then ensues. Elongation of the chordae also causes leaflet tissue to move into the atrium during diastole, also resulting in compromised coaptation. Ruptured or flail leaflets are often the result of systolic stresses fracturing weakened chordae, causing severe regurgitation.

Mitral regurgitation represents pure volume overload for the LV.²² In myxomatous disease, MR is typically of the chronic compensated variety.²³ A vicious cycle is perpetuated whereby the excess volume load over time results in ventricular failure. Ventricular failure itself implies ventricular dilatation, which results in a greater degree of MR. Thus, MR begets more MR, and a downward spiral is created.¹⁷ MR created by left ventricular failure occurs primarily by ventricular dilatation. Successful repair will result in left ventricular mass reduction.²⁴ Ventricular dilatation “pulls” or “tethers” the leaflets open, thereby impinging on coaptation. In degenerative disease of the mitral valve, however, the LV itself per se does not primarily cause MR in the early course of the disease.

Carpentier¹⁰ has developed a mitral valve analysis protocol and surgical philosophy for repair of all types of valves (Fig. 35-4). Myxomatous disease has become the pathology responsible for the vast majority of patients with mitral regurgitation in the United States.²⁵ Currently, mitral valve prolapse, as a part of the spectrum of degenerative disease, is present in about 5% of the general population,²⁶ with about 10% of these patients exhibiting severe MR requiring surgery.²⁷ Whatever the ultimate pathologic pathway creating regurgitation, more than 95% of all degenerative disease should be amenable to successful repair.

Diagnostic Work-up and Indications for Operation

Patient presentation is typically quite varied depending on the degree of MR as well as the chronicity of the disease. Patients may be floridly symptomatic or completely asymptomatic. Symptoms can occur in different forms. Heart failure symptoms are secondary to pulmonary venous hypertension as well as fluid retention. This may include shortness of breath, limited exertional capacity, fluid overload, and in the late stage of the disease, frank heart failure. Embolization sequelae and arrhythmias form a second set of symptoms and include atrial fibrillation and increased stroke risk.²⁸ Additionally, regurgitation predisposes the valve to infectious endocarditis.²⁹ Abnormal hemodynamics create pathologic shear stresses and turbulence that generates vulnerability to infection in the valve.

Evaluating MR

Key information is obtained from echocardiography: the degree of MR, associated pathophysiology, cardiac chamber dimensions, and left ventricular functional analysis. Transthoracic echocardiography is usually the first modality employed. However, if images from transthoracic echocardiography are of insufficient quality, then transesophageal echocardiography (TEE) is required. Mitral regurgitation is graded on a scale from mild to severe, with severe typically referring to a reversal of pulmonary venous blood flow in systole.³⁰ The methods used to determine the degree of regurgitation commonly include regurgitant volume, regurgitant fraction, and effective regurgitant orifice area.³⁰ Echocardiographic analysis of the MR (eg, flail leaflet, ruptured chordae, or anterior or posterior prolapse) is extremely helpful in planning the operative intervention. Other crucial information obtained by the preoperative echo includes left atrial size, ventricular function, ventricular dilatation, aortic valve function, and tricuspid valve function. A large left atrium implies chronic MR. A small LA and hyperdynamic LV implies acute MR. Ventricular function is a key component in assessing operative candidacy; a left ventricular ejection fraction below the norm of 60% indicates some degree of myocardial decompensation secondary to volume overload. Indeed, in the presence of severe MR, the ejection fraction will often decline postoperatively even if the preop ejection fraction is normal. A corollary of this fact is a normal preoperative ejection fraction does not necessarily mean normal ventricular function.

Timing of Surgery

Once pulmonary hypertension, LV dysfunction/LA dilatation or symptoms appear, the prognosis of MR worsens. Careful surveillance may result in timing of valve surgery before these negative sequelae occur.^30a An attractive alternative strategy for treating severe chronic primary MR is to perform early mitral repair before these triggers are reached. Early mitral repair avoids the need for intensive surveillance and also obviates the possibility that patients might become lost to follow-up or delay seeing their clinician until advanced LV dysfunction has already ensued. This strategy requires expertise in clinical evaluation and cardiac imaging to evaluate severity and significance of MR.

What, precisely, are the indications for operation? In general, as surgical results improve, the indications for surgery are broadening. The American Heart Association 2014 guidelines state that “Mitral valve repair is reasonable in asymptomatic patients with chronic severe primary MR (stage C1) with preserved LV function (LVEF > 60% and LVESD 0 40 mm) in whom the likelihood of a successful and durable repair without residual MR is greater than 95% with an expected mortality rate of less than 1% when performed at a Heart Valve Center of Excellence.” Patients who would never have been considered for surgery previously are now routinely being offered repair surgery at much earlier stages of their disease. The indications for mitral valve repair have widened as the success of repair has so dramatically improved over the last several years. Better myocardial protection and cardiopulmonary bypass technology, minimally invasive incisions, increased incidence of repair, and better intensive care unit support have all contributed to the phenomenon.³¹ The biggest change has been an overall broadening of the indications for mitral valve repair and lowering of the threshold for operation because of these factors,³² even in the elderly.³³ Increasingly, asymptomatic mitral regurgitation, even without symptoms, is becoming accepted as a reasonable strategy.³⁴ Repair of the mitral valve, as opposed to replacement, is now accepted as the superior treatment for myxomatous mitral regurgitation. Out of a long laboratory and clinical experience have come the conclusions that repair is associated with better survival, enhanced preservation of ventricular function (by preserving the chordae and papillary muscles), and decreased late thromboembolic complications.^35-42

For all patients with moderate-to-severe MR and moderately decreased ventricular function (left ventricular ejection fraction < 60%) valve repair is offered, because the ventricle has exhibited signs of decompensation even with a lesser degree of MR; repair in this setting is much more urgent, because severe decompensation can occur in a matter of months as left ventricular function begins to deteriorate.

The standard of care for patients with severe cardiomyopathy and severe MR is unclear presently and is one of the most controversial topics in cardiac surgery. This situation typically does not occur with myxomatous disease but rather ischemia or idiopathic cardiomyopathy and such regurgitation is functional in nature. In longstanding myxomatous disease, however, LV dysfunction may be present.

The intermediate situation of moderate mitral regurgitation and preserved ventricular function is the one in which the most judgment must be exercised. Consider the situation of a structurally normal mitral valve with moderate functional MR, but with concomitant critical aortic stenosis. The patient is to undergo an aortic valve replacement. Should the mitral valve be repaired? Perhaps not, as the moderate MR here is typically exacerbated by the patient’s aortic stenosis and volume overload, and correction of the aortic stenosis and euvoluemia would likely reduce or eliminate the MR in this structurally normal valve.^39,40 However, should repair be undertaken, ring annuloplasty would be needed to correct the posterior annular dilatation. If the same situation exists, however, with a structurally abnormal valve, such as a prolapsed P2 or markedly dilated mitral annulus in a myxomatous valve, then mitral repair should always be undertaken, because there is a structural abnormality that aortic valve replacement will not correct. What of moderate MR secondary to myxomatous disease (and not ischemia) in a patient who is to undergo coronary bypass grafting? That patient would likely be offered concomitant valve repair, because bypass grafting would not impact on the pathophysiology of the MR in this case. What of the situation of moderate-to-severe isolated MR secondary to myxomatous disease and borderline normal ventricular function? Even without symptoms of heart failure, this patient should have repair performed on an elective basis. Once left ventricular function begins to deteriorate from the normal 60 to 70% left ventricular ejection fraction, decline may be unpredictably rapid and hence intervention warranted. Increasingly, moderate ischemic, as well as degenerative, MR is thought to be detrimental for long-term survival.^41,42

In general, no matter what the chronological age is, adequate functional status before valve repair is preferred. If symptoms of heart failure exist before surgery, optimal diuresis should be undertaken before operation. If the procedure involves coronary artery bypass grafting, the conduit status should be determined. Dental clearance on all patients should be obtained before any valvular procedure. If neurologic symptoms exist or if the patient has a history of a previous cerebrovascular disease, then preoperative carotid noninvasive studies are warranted to assess carotid arterial stenosis. All patients older than 40 years should undergo coronary angiography. Patients with MR without coronary artery disease are good candidates for smaller incisions.^43,45

Whatever the scenario, the decision to operate and repair the valve is a decision made preoperatively as opposed to intraoperatively. Once the patient is under anesthesia, loading conditions are not physiologic and the mitral regurgitation assessed then is inevitably underestimated. Maneuvers used to “bring out the MR,” such as increasing afterload with vasoactive drugs, do not reflect true physiology, but should be used in surgical decision making and may be helpful in some situations. Recent discussions of earlier referral for the treatment of MR clearly depend on a high rate of valve repair in any center.^46,47

Operative Philosophy

Clearly, in light of elevated success rates of mitral valve repair in specialized centers but also today in all cardiac surgery programs, the patient with degenerative mitral valve disease with significant MR requires a repair procedure, rather than replacement.⁴⁹

With the increasing numbers of trained surgeons who have absorbed the various mitral valve repair techniques and can apply them safely the percentage of valves that can be repaired and actually are repaired, is improving. Recent STS database interrogation indicates that this percentage may now be as high as 75%.⁴⁹

Beginning in the 1980s, we have developed what we think is a simple, reproducible algorithm that can be used to repair the degenerative mitral valve in most patients. Overriding this philosophy is the belief that mitral valve repair is not an esoteric “art form” that is difficult to explain, perform, or learn. Rather, we think it is a procedure like any other that should and can be simplified, disseminated, and reproduced with success. In keeping with this philosophy, we have reduced complicated bileaflet prolapse to a competent valve with simplified and straightforward maneuvers that we have found effective with good long-term results. Our overall philosophy and technique are as follows:

1. 
   

   Expose the valve well through the complete development of Sondergaard’s groove, division of pericardial attachments of the superior and inferior venae cavae, and release of the left pericardial retraction stitches

   
   
2. 
   

   Assess the valve through saline injection and corroborate the intraoperative findings with TEE

   
   
3. 
   

   Perform basic, obvious leaflet repair procedures first (eg, quadrangular resections to the posterior leaflet)

   
   
4. 
   

   Implant the annuloplasty ring sized by the height of the anterior leaflet (not the trigones or commissures)

   
   
5. 
   

   Test the repair

   
   
6. 
   

   Perform additional reparative procedures as needed, that is, cleft closure.

With the above maneuvers, we estimate that approximately 95% of all degenerative valves can be repaired.

Collaborative involvement with cardiac anesthesia colleagues is essential in utilizing the invaluable tool of transesophageal echocardiogram monitoring for pre- and postrepair assessment. In our clinic, standard TEE monitoring (either two- or three-dimensional) is now utilized for every patient undergoing mitral valve repair. In addition to documenting the efficacy of repair, TEE is essential in preventing and assessing the potential or persistence of systolic anterior motion (SAM) of the anterior valve.

Operative Exposure

Because the mitral valve is such a complex anatomic structure and the maneuvers involved in correcting a regurgitant valve may vary from the simple to the very complex, adequate exposure is an absolute requirement in every operative plan. This becomes more important if minimally invasive techniques are employed. Mitral valve exposure is more challenging than exposure of the aortic or tricuspid valve. Why? From the surgeon’s side, the mitral valve is furthest away from the operating surgeon than any other valve. In addition, the valve in its native position naturally faces above and/or toward the surgeon’s left shoulder at an oblique angle such that the surgeon does not see the valve en face.

The first critical aspect to the standard valve repair is the complete and thorough development of the Sondergaard plane reflecting the right atrium off the left atrium to the atrial septum, as depicted in Fig. 35-5. This was first described in the 1950s by the Danish surgeon Sondergaard,⁵⁰ to expose the atrial septum for noncardiopulmonary bypass treatment of atrial septal defects. In 1990, we stressed the importance of this particular technique for exposure in mitral valve surgery.⁵¹ Regardless of even previous procedures, it should always be possible to dissect out the groove without significant difficulty. The complete and full development of the groove is a most important aspect to obtain adequate exposure of the mitral valve. With this technique, via blunt and sharp dissection we typically have not needed any other incision for mitral valve repair or replacement, whether for primary surgery or reoperation. This incision usually brings the surgeon very close to the mitral valve. Once the right atrium is dissected off the left atrium, a generous incision in the left atrium is made, avoiding the atrial septum.

The second aspect of exposure is to bring the valve as close to an en face position to the operating surgeon as possible. Pericardial stay sutures are released on the left side. The operating table is maneuvered head up and to the left. If necessary, the left pericardium is opened and the apex of the heart moved laterally to the left pleural space. Exposure of the left trigone can be particularly challenging for stitch placement. This can be facilitated with local epicardial displacement of the midlateral LV wall medially with aspongestick (Fig. 35-6).

Cardiopulmonary Bypass

For cardiopulmonary bypass, we use a 22F percutaneous femoral vein venous catheter placed into the right atrium via the right femoral vein with TEE control. The cannula can even be advanced into the superior vena cava if desired. Because the cannula has multiple holes, is flexible and thus can still drain the inferior vena cava, this one cannula is sometimes all that is needed for drainage of both the superior and inferior vena cava. If performing a concomitant bypass procedure requiring a full sternotomy, then venous cannulation should be bicaval via the atria. One venous cannula should be placed in the superior vena cava above the right atrial/superior vena caval junction and the other through the lowest part of the right atrium into the mouth of the inferior vena cava. The arterial cannula should be placed directly into the distal ascending aorta.

Once on cardiopulmonary bypass, systemic temperature is allowed to drift to 34°, the ascending aorta is cross-clamped, and the heart arrested by cold blood cardioplegia. For isolated valve repair, some debate still exists regarding the use of retrograde or antegrade blood cardioplegia after cross-clamping. If there is concomitant coronary artery disease, myocardial protection in this circumstance should be antegrade and retrograde, because the coronary artery bypass graft/mitral operation presents one of the highest-risk operative settings in cardiac surgery.³⁶

After the heart is arrested, the left atrium is opened well above the right superior pulmonary vein near the septum inferiorly. Retractors are placed (Fig. 35-7). The patient’s bed is placed head up with a tilt to the left. A wire-reinforced suction catheter is placed in the left inferior pulmonary vein (the most dependent portion of the left atrium in this position) for drainage of collateral blood flow. Carbon dioxide is infused to minimize intracardiac air.