Chapter 42. Mitral Valve Replacement

This chapter discusses the surgical indications, operative techniques, and early and late follow-up after implantation of mechanical and bioprosthetic mitral valve devices. The valves that are discussed are those that are currently (2010) approved by the FDA. Figure 42-1 shows the current FDA-approved prosthetic mitral valve devices, including the Starr-Edwards ball-and-cage valve (historical relevance only), the Omnicarbon tilting-disk valve, the Medtronic Hall tilting-disk valve, the St. Jude Medical bileaflet valve, the Carbomedics bileaflet valve, the ATS bileaflet valve, and the On-X bileaflet valve. The FDA-approved bioprosthetic valve devices are shown in Fig. 42-2 and include the Hancock II porcine valve, the Carpentier-Edwards porcine valve, the Carpentier-Edwards pericardial valve, the Mosaic porcine valve, and the Biocor porcine valve.

Heart valve prostheses are continually undergoing iterative advancement by the manufacturers, and the future device of choice has yet to be developed. This ideal replacement valve prosthetic would have longevity of a mechanical prosthetic combined with the superior hemodynamic function of the native biologic tissue valve. As a result this hypostatic ideal replacement device would not require anticoagulation and carrying no risk of either thromboembolic events or valve thrombosis. To achieve this goal will require major advancement away from current design.

The indications for mitral valve replacement are variable and undergoing evolution. Because of increasing use of reparative techniques, particularly for mitral regurgitation, replacement or repair of a mitral valve often depends on the experience of the operating surgeon. Current indications for valve replacement pertain to those types of valve problems that are unlikely to be repaired by most surgeons or which have been shown to have poor long-term success after reconstruction. Indications are discussed according to: (1) pathophysiologic states; and (2) type of valve required (ie, mechanical or bioprosthetic).

Mitral stenosis is almost exclusively caused by rheumatic fever, even though a definite clinical history can be obtained in only about 50% of patients. The incidence of mitral stenosis has decreased substantially in the United States in the last several decades because of effective prophylaxis of rheumatic fever, nevertheless in certain developing countries mitral stenosis is still very common. Two-thirds of patients with rheumatic mitral stenosis are female.

The pathologic changes associated with rheumatic valvulitis are mainly fusion of the valve leaflets at the commissures, shortening and fusion of the cordae tendineae, and thickening of the leaflets owing to fibrosis with subsequent stiffening, contraction, and calcification. Approximately 25% of patients have pure mitral stenosis, but an additional 40% have combined mitral stenosis and mitral regurgitation.1

Stenosis usually develops one or two decades after the acute illness of rheumatic fever with no or slow onset of symptoms until the stenosis becomes more severe. Limitation of exercise tolerance usually is the first symptom, followed by dyspnea that can progress to pulmonary edema. New-onset atrial fibrillation and the risk for thromboembolism, hemoptysis, and pulmonary hypertension are other common symptoms in patients with mitral stenosis.

The diagnostic workup of the symptomatic patient with mitral stenosis should include a complete cardiac catheterization, including coronary angiography in any patient greater than age of 40. Under age 40, echocardiographic findings of the mitral valve suffice in most symptomatic patients for the definition of mitral valve pathology unless there is a history of chest pain or coronary artery disease. Cardiac catheterization establishes the extent of mitral valve stenosis by determining valve gradients and valve area. Pulmonary artery pressure, which may be extremely high in longstanding cases of mitral stenosis, is also documented. In general, operation is prescribed when the mean valve area is 1.0 cm2 or less2 (normal mitral valve area is 4 to 6 cm2); however, with a “mixed” lesion of mitral stenosis and mitral regurgitation, the valve area in symptomatic patients occasionally may be as large as 1.5 cm2. Asymptomatic patients generally are not considered for surgery,1 but some authors recommend operation in asymptomatic patients with significant hemodynamic mitral stensosis.2 The degree of pulmonary artery pressure elevation secondary to mitral stenosis continues to be an area of concern for the mitral valve surgeon. There is still no definitive answer to this question, but most surgeons operate on patients with severe pulmonary hypertension (suprasystemic patients) with the knowledge that intensive postoperative respiratory and diuretic therapy is necessary to maintain relatively dry lungs and reduce the risk of severe right ventricular failure. It has been known for more than 40 years that after mitral valve replacement for mitral stenosis, pulmonary artery pressure decreases within hours in most patients and decreases more gradually over weeks to months in others.4–6

Success with closed commissurotomy after World War II and the development of the Starr-Edwards valve in the early 1960s led to an enormous increase in operations for rheumatic mitral valve disease. In the 1990s, balloon dilation of fibrotic, stenotic mitral valves became increasingly common.6,7 At the present time, percutaneous mitral balloon valve dilation is used in most cases of symptomatic noncalcified, fibrotic mitral stenosis. But even though this technique has been shown to be equivalent in the short run to closed mitral commissurotomy, especially in young patients, it is only indicated in a minority of patients, that is, those with optimal valvular characteristiscs.1,8 Open mitral commissurotomy and valvuloplasty for such patients can be a successful operation,9,10 but other studies have shown better long-term results with mitral valve replacement using a mechanical valve.11 Many patients with chronic mitral stenosis now require valve replacement because the valve has developed significant dystrophic changes, including marked thickening and shortening of all chordae, obliteration of the subvalvular space, agglutination of the papillary muscles, and calcification in both annular and leaflet tissue. Aggressive decalcification and heroic reconstructive techniques for these extremely advanced pathologic valves generally have produced poor long-term results; nevertheless, some surgeons still advocate aggressive repairs in this subset of patients.12

The etiology of mitral regurgitation is very diverse, and the decision to recommend operation for patients with mitral regurgitation is more complex than for patients with mitral stenosis, except in patients with acute ischemic mitral regurgitation and endocarditis, in whom indications are more straightforward. The pathologic subsets that produce mitral regurgitation are related to a number of metabolic, functional, and anatomical abnormalities.1 These can be categorized into degenerative (eg, mitral prolapse and ruptured/elongated chordae), rheumatic, infectious, and ischemic diseases of the mitral valve. Most of these entities are now amenable to mitral valve repair and reconstruction with or without the use of annuloplasty rings, as mentioned in Chapter 41.

It is important to stress that depressed ejection fraction is a poor indicator of left ventricular function in patients with mitral regurgitation. Ejection fraction can be preserved in patients with irreversible left ventricular failure because of regurgitant flow through the valve.20,21 Depressed cardiac output (<40%) therefore usually indicates severe left ventricular dysfunction, and results of surgery are not as favorable in these patients as they are in those with normal ventricles.22,23 Compared with ejection fraction, measurements of end-systolic volume and diameter are more reliable noninvasive parameters to evaluate the status of the left ventricle and determine the optimal time for operation24,25

Once the valve is exposed, indications for mitral valve replacement in patients with mitral regurgitation depend on the extent of the pathology in each patient and the reparative experience of the operating surgeon. Thus, in regurgitation from degenerative prolapsing myxomatous valves that have a high probability of reconstruction, mitral valve repair is indicated if the prolapse is generalized and local findings that decrease the probability of a successful repair are absent.26–30 Similarly, if rheumatic mitral regurgitation, calcific deposits throughout the leaflet substance, and shortened chordae and papillary muscles are encountered, mitral valve replacement is often the most prudent operation because the probability of successful repair is low.31 However, good results with reconstructive surgery in this patient group have been reported.32 In ischemic mitral regurgitation, pathology that precludes satisfactory repair includes restrictive valve motion from shortened, scarred papillary muscles, an acutely infarcted papillary muscle, and rupture of chordae associated with extensive calcification of valve leaflets.33–35 In endocarditis, mitral valve replacement may be required because of destruction of the valve leaflets and subvalvular mechanisms and annular abscess formation. Although repair of the valve and avoidance of prosthetic material are very desirable in septic situations, the extent of the destruction may preclude repair. Therefore, mitral valve replacement is required after careful debridement of the infectious tissue and reconstruction of the valve annulus.36–38

Indications for Mechanical Valve Replacement

Currently available prosthetic valves in the United States are the bileaflet and the tilting disk designs. For young patients, patients in chronic atrial fibrillation who require long-term anticoagulation, and any patient who wants to minimize the chance of reoperation, a prosthetic valve should be chosen if valve replacement is required. The St. Jude Medical bileaflet valve is the most widely used prosthetic mitral valve at present because it has good hemodynamic characteristics and is easy to insert. Recent interest in the On-X mechanical valve relates primarily to the possibility for limited anticoagulation, although this remains to be borne out from current clinical trial.39 Indications to choose one prosthetic or another vary primarily by surgeon preference and occasionally depending on the state of the annulus and whether or not there have been multiple previous operations. For example, infrequently, the mitral annulus provides poor anchorage with subsequent perivalvular leak with the bileaflet or tilting-disk valve, which requires an everting-suture technique. In this instance, a valve with a bulky sewing ring may be chosen to reduce the probability of subsequent perivalvular leak. A low-profile mechanical valve is preferable in a patient with a small left ventricular cavity to prevent obstruction of left ventricular outflow and impingement of the myocardium.

Indications for Bioprosthetic Valve Replacement

Patients in any age group in sinus rhythm who wish to avoid anticoagulation may prefer a bioprosthetic valve. This is especially true for patients in whom anticoagulation is contraindicated; for example, patients with a history of gastrointestinal bleeding or those who have a high-risk occupation or lifestyle.1 A bioprosthetic valve is preferred in patients greater than age 65 and in sinus rhythm because these valves deteriorate more slowly in older patients.40In addition, some 60-year-old patients may not outlive their prosthetic valves because of comorbid disease.41,42 Specifically, patients who require combined mitral valve replacement and coronary bypass grafting for ischemic mitral regurgitation and coronary artery disease have significantly reduced long-term survival as compared with patients who do not have concomitant coronary artery disease.43–48 These individuals may avoid anticoagulation with little risk of reoperation.

As 20-year results have become available for various bioprostheses, it is clear that structural valve degeneration (SVD) is the most prominent complication of these valves.49–54 The durability of porcine mitral valves is less than with aortic bioprostheses. The more rapid deterioration of mitral bioprostheses may be caused by higher ventricular systolic pressures against the mitral cusps compared with the diastolic pressures resisted by aortic bioprosthetic leaflets. Durability of bioprosthetic valves is directly proportional to age50; deterioration occurs within months or years in children and young adults and only gradually over years in septuagenarians and octogenarians.49,55 Essentially all valves implanted into patients younger than 60 years of age have to be replaced ultimately, and valve failure is prohibitively rapid in children and adults younger than 35 to 40 years of age; therefore, bioprostheses are not advisable in these age groups.56 Nevertheless, there are still indications for mitral porcine bioprosthetic valves in young patients. In a woman who desires to become pregnant, a bioprosthesis may be used to avoid warfarin anticoagulation and fetal damage during pregnancy.57–60 In patients with chronic renal failure and hypercalcemia related to hyperparathyroidism, bioprostheses have extremely limited durability and therefore should be avoided.

Over the last decade several reports, mainly from European centers, on the use of unstented cryopreserved homografts61–64 and stentless heterografts65–68 for mitral valve replacements, particularly in patients with endocarditis. The prosthetic valve is transplanted, donor papillary muscles are reattached to recipient papillary muscles, and the annulus is sutured circumferentially. This technique has been shown to be safe and reproducible, but it does not always provide durable results and therefore should not be used in young patients.65 Other reports suggest that these operations may be a feasible alternative to stented valve replacement in patients with endocarditis. Pulmonary autografts also have been used for replacing the mitral valve (Ross II procedure), but these series are small, and follow-up is relatively short.69–71

Mitral valve surgery has been in a state of constant evolution since its inception. Data regarding all cardiac surgical procedures reported to the Society of Thoracic Surgeons Adult Cardiac Surgery Database (STS ACSD) demonstrate this dynamic state as more surgeons begin to repair rather than replace mitral valves. Gammie et al. recently evaluated trends in mitral valve surgery in the United States using the STS ACSD evaluating the years between 2000 and 2007.192 In this time period, 210,529 mitral valve procedures were performed in all settings. From the study population 58,370 patients undergoing primary mitral valve operations were identified. Over this 7-year study time line, a 50% increase in repair rates was documented. When considering the valve procedures involving replacement, a 100% increase in use of bioprosthetic devices coincided with a concomitant reduction in use of mechanical valves.

Gammie and coworkers identified clear trends with respect to patient selection as well.192 Compared with patients undergoing mitral valve repair, those undergoing mitral replacement tended to be older, female, more likely to have multiple comorbidities (eg, diabetes mellitus, hypertension, chronic lung disease, stroke), concomitant tricuspid valve disease, and mitral stenosis, and were less likely to be asymptomatic. With respect to survival, overall risk adjusted mortality was lower for mitral valve repair versus replacement (OR 0.52, 95% CI: 0.45 to 0.59, p < .0001). The outcome with respect to choice of valve type and survival was not addressed but was felt to be confounded by patient factors that led to the initial device selection.

Needless to say, the role for mitral valve replacement as a whole is being diminished by improvements in mitral repair technique. Most importantly, these technical advancements have simplified repair procedures such that more surgeons can perform the procedure while not compromising on repair results. Each individual surgeon must assess their own outcomes to judge whether their ability to repair the valve can offer the same if not superior results when compared with replacement in their hands.

Mechanical Prosthesis

The designs of mechanical and bioprosthetic heart valves have evolved over the last five decades in an effort to develop the ideal replacement for the pathologic mitral valve. Biochemical and engineering advances have produced hemodynamic improvements and reduced morbidity from valve-related complications. The ideal valve, however, is not available, and the positive and negative characteristics of current valves must be considered when choosing the most appropriate valve for an individual patient. The optimal heart valve exerts minimal resistance to forward blood flow and allows only trivial regurgitant backflow as the occluder closes. The design must cause minimal turbulence and stasis in vivo during physiologic flow conditions. The valve must be durable enough to last a lifetime and must be constructed of biomaterials that are nonantigenic, nontoxic, nonimmunogenic, nondegradable, and noncarcinogenic. The valve also must have a low incidence of thromboembolism.

The opening resistance to blood flow is determined by the orifice diameter; the size, shape, and weight of the occluder; the opening angle; and the orientation of leaflet or disk occluders with respect to the plane of the mitral annular orifice for any given annular size. Least resistance to transvalvular blood flow during diastole for valves in the mitral position is provided by a large ratio of orifice to total annular area. A wide opening angle also improves the effective orifice area and results in decreased diastolic pressure gradients. With an increasing orifice diameter, however, more energy is lost across the valve as more backflow passes through the valve at end diastole and early systole. Table 42-1 shows hemodynamic assessments of each of the FDA-approved mitral valve prostheses for the most commonly used mitral valve sizes.72–75 The results of in vivo assessments at rest by invasive (catheterization) or noninvasive (Doppler echocardiography) techniques are tabulated.

Blood turbulence flowing across mitral valve devices results from impedance to forward or reverse flow. This impedance can be minimized by occluder design and orientation, central flow through the orifice, and limited struts or pivots extending into flow areas (Fig. 42-3). Hemolysis is the product of red blood cell destruction that is caused by cavitation and shearing stresses of turbulence, high-velocity flow, regurgitation, and mechanical damage during valve closure.76 Areas of perivalvular blood stagnation and turbulence increase platelet aggregation, activation of the coagulation proteins, and thrombus formation.

Dynamic regurgitation is a feature of all prosthetic valves and is the sum of the closing volume during occluder closure and the leakage volume that passes through the valve while it is closed. The closing volume is a function of the effective orifice area and the time needed for closure. Closure time is influenced by the difference between the opening and closing angles of the occluder and valve ring. Leakage volume is inherent to the design of the valve and depends on the amount of time the valve remains in the closed position.77 A small amount of regurgitant volume can be beneficial by minimizing stasis and reducing platelet aggregation; this decreases the incidence of valve thrombosis and valve-related thromboembolism.

The Starr-Edwards Model 6120 is the only ball-and-cage mitral valve prosthesis currently approved for use in the United States by the FDA. It was introduced with its current design in 1965 after undergoing several engineering modifications and has been in use longer than any other type of mechanical valve (see Fig. 42-1A). The occluder is a barium-impregnated Silastic ball in a Stellite alloy cage that projects into the left ventricle. This valve has a large Teflon/polypropylene sewing ring that produces a relatively smaller effective orifice and larger diastolic pressure gradients than other prosthetic valves of similar annular sizes. Leakage volumes are not inherent in the ball-and-cage design, and in contrast with other mechanical valves, the presence of regurgitation may indicate a pathologic process. The central ball occluder causes lateralization of forward flow and results in turbulence and cavitation that increase the risk of hemolysis and thromboembolic complications (see Fig. 42-3A). The incidence of thromboembolism has been shown to be higher with the Starr-Edwards valve than with bileaflet valves in most studies78,79 but not all.80 Because the cage projects into the left ventricle, it is unwise to implant this valve in small left ventricles, where the cage may contact the ventricular wall or cause ventricular outflow obstruction.

Tilting-disk mitral valve prostheses have better hemodynamic characteristics than ball-and-cage valves (see Fig. 42-3B). The Medtronic Hall central pivoting-disk valve was introduced in 1977 and is based on engineering design modifications of the earlier Hall-Kaster valve81 (see Fig. 42-1B). The axis of the tilting disk was moved more centrally to allow greater blood flow through the minor orifice and to reduce stagnation in areas of low flow. The opening angle originally was increased to 78 degrees to decrease resistance to forward flow and later narrowed to 70 degrees when in vitro studies revealed an unacceptable regurgitant volume. The opening angle of 70 degrees produced regurgitation volumes of less than 5% of left ventricular stroke volume without significantly compromising forward flow. The disk occluder was allowed to slide out of the housing at the end of the closing cycle to provide a gap through which blood could flow to minimize stasis at the contact surfaces.110 The large opening angle and slim disk occluder, along with a thinner sewing ring, provide improved hemodynamics with comparably larger effective orifice areas and lower mean diastolic pressure gradients for each valve size. During implantation, the larger orifice should be oriented posteriorly when using the larger valve sizes to minimize the potential for disk impingement. Smaller valves (27 mm or less) should be oriented with the larger orifice anteriorly to optimize in vivo hemodynamics.73,82

The Omniscience tilting-disk valve is a second-generation device derived from improvements to the design of the Lillehei-Kaster pivoting-disk valve.83 This low-profile device has a pyrolytic disk located eccentrically in a one-piece titanium housing attached to a seamless Teflon sewing ring. Introduced in 1978, the Omniscience prosthesis includes several engineering modifications from prior devices in an effort to improve its hemodynamic function. The orifice-to-annular-area ratio was increased to minimize resistance to forward flow. The opening angle of 80 degrees is relatively large to allow flow reserve in patients with high cardiac outputs and during exercise. Resulting increases in regurgitant volumes are minimized by the disk design. Turbulence is reduced by the curvature of the disk, and areas of stasis and shear stress are reduced by the eccentric location of the pivot axis in an effort to decrease the risk of thrombosis, thromboembolism, and hemolysis. Retaining prongs are not used, and the lower profile reduces the risk of impingement. A potential hemodynamic disadvantage that has been the subject of debate is the possibility of incomplete disk opening in vivo. Clinical studies report postoperative mean opening angles of between 44.873 and 75.9 degrees.84 Implicated factors causing this variation include valve sizing, orientation during implantation, and anticoagulation status.84,85 A subsequent generation of the Omniscience valve is the all-carbon Omnicarbon monoleaflet valve that was released in 2001 in the United States but has been in clinical use in Europe since 1984 (see Fig. 42-1C). The housing material is made of pyrolytic carbon instead of titanium. As a result of this change, the incidence of thromboembolism, valvular thrombosis, and reoperations was decreased significantly compared with the Omniscience valve protheses.86 For all the tilting-disk valves, meticulous surgical technique is important because retained leaflets or chordae can cause subvalvular interference and leakage.

The unique design of the bileaflet St. Jude Medical valve was introduced in 1977, and it is currently the prosthesis used most commonly worldwide (see Fig. 42-1D). Two separate pyrolytic carbon semidisks in a pyrolytic carbon housing are attached to a Dacron sewing ring. The housing has two pivot guards that project into the left atrium. The bileaflet design produces three different flow areas through the valve orifice that provide overall a more uniform, central, and laminar flow than in the caged-ball and monoleaflet tilting-disk designs. The improved flow results in less turbulence and decreased transmitral diastolic pressure gradients77,87 (see Fig. 42-3C) at any annulus diameter size and cardiac output compared with the caged-ball and single-leaflet tilting valves.88 The favorable hemodynamics in smaller sizes makes it especially useful in children. The central opening angle is 85 degrees, with a closing angle of 30 to 35 degrees, which along with a thin sewing ring, provides a large effective orifice area for each valve size at the expense of greater regurgitant volumes, especially at low heart rates. Asynchronous closure of the valve leaflets in vivo also contributes to the regurgitant volume.89 The design of this prosthesis provides excellent hemodynamic function even in small sizes in any rotational plane.90 The antianatomical plane, however, with the central slit between the leaflets oriented perpendicular to the opening axis of the native valve leaflets decreases the potential risk of leaflet impingement by the posterior left ventricular wall.91

The Carbomedics bileaflet valve was approved by the FDA in 1986 (see Fig. 42-1E). This low-profile device is constructed of pyrolytic carbon and has no pivot guards, struts, or orifice projections to decrease blood flow impedance and turbulence through the valve.77 It has a rotatable sewing cuff design and is available with a more generous and flexible sewing cuff (the OptiForm variant) that confirms more easily to different patient anatomies and allows sub-, intra-, or supra-annular suture placement. The leaflet opening angle is 78 degrees, which with the bileaflet design provides a relatively large effective orifice area and transvalvular diastolic pressure differences only slightly greater than the St. Jude Medical bileaflet valve. Rapid synchronous leaflet closure reduces closing regurgitant volumes to less than that of the Björk-Shiley pivoting-disk prosthesis, which has an opening angle of 60 degrees. Leakage volume, however, is greater with Carbomedics valves because of backflow through gaps around pivots. Because of its narrow closing angle and large leakage volume, the Carbomedics valve does not reduce the relatively large regurgitant volume associated with the bileaflet design. Although this valve has good hemodynamic function overall, in the mitral position, the 25-mm Carbomedics valve has a relatively high diastolic pressure gradient and large regurgitant energy loss across the valve, especially at high flows. Hemodynamic studies suggest that the Carbomedics valve should be avoided in patients with a small mitral valve orifice.77

The ATS (Advancing The Standard) mechanical prosthesis has been in clinical use in the United States since 2000. Similar to the Carbomedics valve, the ATS valve is a low-profile bileaflet prosthesis with a pyrolytic housing and pyrolytic carbon leaflets containing graphite substrate (see Fig. 42-1F). The pivot areas are located entirely within the orifice ring, and the valve leaflets hinge on convex pivot guides on the carbon orifice ring. This design minimizes the overall height of the valve and provides a wider orifice area, and the absence of cavities in the valve ring theoretically reduces stasis or eddy currents that may develop. Valve noise, a bothersome problem for some patients, also is reduced by this design.91 The opening angle is up to 85 degrees, and the sewing cuff is constructed of double velour polyester fabric that is mounted to a titanium stiffening ring, which enables the surgeon to rotate the valve orifice during and after implantation.

The On-X prosthesis was approved by the FDA in 2002. It has a bileaflet design similar to the St. Jude Medical, Carbomedics, and ATS prostheses with comparable hemodynamic performance, ie, a relatively large orifice diameter and a wide opening angle (90 degrees) (see Fig. 42-1G). Instead of silicon-alloyed pyrolytic carbon, as used in the other mechanical prostheses, the On-X valve is made of pure pyrolytic carbon. This material is stronger and tougher than silicon-alloyed carbon and allows incorporation of hydrodynamically efficient features into the valve orifice, such as increased orifice length and a flared inlet that reduces transvalvular gradient. Early clinical results are promising and the valve produces very little hemolysis with postoperative levels of serum lactate dehydrogenase in the normal range.95,96

Bioprostheses

Porcine Valves

The porcine bioprosthetic mitral valves are designed to mimic the flow characteristics of the in situ aortic valve. The Hancock I mitral valve bioprosthesis was introduced in 1970. It has three glutaraldehyde-preserved porcine aortic valve leaflets on a polypropylene stent attached to a Dacron-covered silicone sewing ring. The design allows for central laminar flow through the valve, which tends to decrease diastolic pressure gradients and minimize turbulence.87 The stent, however, impedes forward flow and results in relatively large diastolic pressure gradients across the bioprosthesis. The stent and the large sewing ring contribute to effective orifice areas that are smaller than those of size-matched mechanical valves (see Table 42-1).

The Hancock II porcine bioprosthesis (see Fig. 42-2A) is the more modern version of the Hancock I prosthesis. The stent is made of Delrin with a scalloped sewing ring and reduced stent profile. The leaflets are fixed in glutaraldehyde at low pressure and subsequently for a prolonged period at high pressure. To retard calcification, the leaflets are treated with sodium dodecyl sulfate.

The Carpentier-Edwards porcine valve uses a flexible stent to decrease the stress of leaflet flexion while maintaining its overall configuration (see Fig. 42-2B). The effective orifice-to-total-annulus-area ratio for the Carpentier-Edwards valve is relatively small, but exercise studies show that the effective orifice area increases significantly with increased blood flow across the valve; diastolic gradients also increase, although to a lesser degree.74,75,97, Porcine bioprostheses in the mitral position should be avoided in patients with small left ventricles because of the possibility of ventricular rupture or left ventricular outflow obstruction caused by the large struts.97

The Mosaic porcine bioprosthesis is a third-generation bioprosthesis using the Hancock II stent (see Fig. 42-2C). It was introduced in the United States in 2000 and has a Delrin stent, scalloped sewing ring, and reduced stent profile. The valve tissue is pressure-free fixed with glutaraldehyde, and the prosthesis is treated with alpha-oleic acid (AOA) to retard calcification.

In 2005, the FDA approved the Biocor porcine bioprosthesis (St. Jude Medical); however, it has been used and investigated for almost two decades in Europe. It belongs to the third generation of bioprostheses, and the valve tissue is pretreated in glutaraldehyde at very low pressure (<1 mm Hg), making the valve cusps less stiff with less tendency to tissue fatigue.

Pericardial Valves

Previous studies indicated poor durability of pericardial valves, namely, the Ionescu-Shiley valve, caused by leaflet tearing. This led to significant changes in design, including mounting of the pericardium completely within the stent, causing less leaflet abrasion and increased durability. The Carpentier-Edwards pericardial valve uses bovine pericardium as material to fabricate a trileaflet valve that is cut, fitted, and sewn onto a flexible Elgiloy wire frame for stress reduction (see Fig. 42-2D). The tissue is preserved with glutaraldehyde with no applied pressure, and the leaflets are treated with the calcium mitigation agent XenoLogiX. Compared with the Carpentier-Edwards porcine bioprosthesis, the stent profile is reduced. Long-term durability for the Carpentier-Edwards pericardial valve is strong, and compared with third-generation porcine valves, valve-related complications are similar (see discussion later in this chapter).

Hemodynamically, pericardial valves provide the best solution to flow problems. The design maximizes use of the flow area, which results in minimal flow resistance. Figure 42-4A shows how the cone shape of the open valve and circular valve orifice minimizes flow disturbance compared with the more irregular cone shape of the porcine valves that allow for central unimpeded flow (see Fig. 42-4B).

Structural valve deterioration is seen after long-term follow-up of patients with both porcine and pericardial bioprostheses and results in mitral stenosis or regurgitation or both. Hemodynamic studies early after operation and at 5 years reveal higher average diastolic pressure gradients and smaller effective orifice areas when compared in the same patients at the follow-up study. In some patients, these changes are sufficiently severe to require reoperation as soon as 4 to 5 years postoperatively, and by 10 years the rate of primary tissue failure averages 30%. It then accelerates, and by 15 years postoperatively, the actuarial freedom from bioprosthetic primary tissue failure has ranged from 35 to 71%49,51,53,54,72 (Table 42-2). Most of these patients show hemodynamic evidence of valvular deterioration before any clinical signs or symptoms.75 Bioprosthetic valves have the advantage of low thrombogenicity, which must be weighed against poor long-term durability and subsequent hemodynamic deterioration and the risk of reoperation.

Preoperative Management and Anesthetic Preparation

Congestive heart failure secondary to mitral stenosis usually can be treated with aggressive diuretic therapy and sodium restriction preoperatively. If the patient is in rapid atrial fibrillation, digoxin, beta blockers, and calcium channel antagonists can be used to slow down the ventricular rate. Patients with acute mitral regurgitation often are in cardiogenic shock, and they can be stabilized preoperatively with inotropes and arterial vasodilators to reduce systemic afterload. Intra-aortic balloon counterpulsation also can be used for this purpose. Symptoms of congestive heart failure in patients with chronic mitral regurgitation are treated with diuretics and oral vasodilators. The vasodilators lower the peripheral vascular resistance, and forward cardiac output is increased by reducing the regurgitant volume into the left atrium.

Preferred anesthesia for mitral valve replacement typically involves a combination of narcotic and inhalational agents. Ultimately, anesthetic management is dictated by the wide range of functional disabilities and hemodynamic abnormalities of patients who present for mitral valve replacement. For example, a cachectic patient with functional class IV mitral stenosis and severe pulmonary hypertension may require postoperative positive-pressure mechanical ventilation for 1 or 2 days to remove excess pulmonary fluid by diuresis, facilitate bronchial toileting, and provide optimal conditions for adequate gas exchange. Alternatively, young patients who require mitral valve surgery and present with less preoperative comorbidity may benefit from a short-acting, balanced anesthetic that can facilitate extubation within 6 hours of surgery.98

Monitoring should include arterial and venous lines, a urinary catheter, and a pulmonary artery catheter placed before bypass to measure pulmonary pressures and cardiac output. After valve replacement, occasionally a left atrial catheter directly inserted through the left atrial incision can be helpful to allow measurement of pulmonary vascular resistance, but we do not use it routinely. Preoperative intravenous prophylactic antibiotics are administered to all patients and are continued for 2 postoperative days until lines are removed. Temporary ventricular pacing wires are placed, and in many instances temporary atrial pacing wires are placed for possible pacing or diagnosis of various atrial arrhythmias.

Cardiopulmonary Bypass for Mitral Valve Replacement

Cardiopulmonary bypass is instituted by placing two right-angle cannulas into the superior and inferior venae cavae. We place a small (22 French) plastic or metal cannula directly into the superior vena cava, above the sinoatrial node. The inferior caval cannula is placed at the entrance of the inferior vena cava, low in the right atrium. These insertion sites keep the caval catheters out of the operative field and yet maintain excellent bicaval drainage. An arterial cannula is placed in the distal ascending aorta. Bypass flows are approximately 1.5 L/m2 per minute, and moderate hypothermia 28 to 32°C is used with vacuum-assisted suction. Myocardial protection includes antegrade and retrograde blood cardioplegia and profound myocardial hypothermia.99 Retrograde cardioplegia is useful for all valve surgery to protect the ischemic left ventricle and help remove ascending aorta bubbles. Antegrade cardioplegia, used as an initial loading dose, is augmented by intermittent retrograde cardioplegia every 20 minutes. This provides safer delivery of cardioplegia because when the atrium is retracted during valve replacement, the aortic valve is distorted, and antegrade cardioplegia tends to fill the ventricle.

Exposure of the Mitral Valve

Evolution of meticulous and complicated methods of mitral valve repair and reconstruction has required optimal exposure of the mitral valve. In primary operations, median sternotomy, development of Sondergaard's plane, and incision of the left atrium close to the atrial septum provide excellent exposure100,101 (Fig. 42-5). This incision is a ubiquitous one, and we have rarely seen indications for use of other incisions, such as the superior approach through the dome of the left atrium,102,103 the so-called biatrial incision popularized by Guiraudon and colleagues,105 division of the superior vena cava,106,107 and the less common but occasionally useful trans–right atrial septal incision.108 The trans–right atrial incision has in some studies been related to higher incidence of junctional and nonsinus rhythm postoperatively,109 although this has not been confirmed by other studies.110

Minimally Invasive Mitral Valve Replacement

After advances in videoscopic and other minimally invasive techniques in many areas of surgery in the 1990s, similar techniques are now being used increasingly in cardiac surgery, especially in mitral valve surgery. In 1996, we began minimally invasive valve surgery for patients who have isolated valvular pathology without concomitant coronary artery disease. Our experience at Brigham and Women's Hospital now totals over 1000 patients, including mitral valve repairs and aortic and mitral valve replacements.110 The minimally invasive approach for mitral valve surgery is well described in Chapter 44.

Safety and efficacy of minimally invasive mitral valve surgery have been confirmed in several reports.31,109–111 Trauma seems to be less with the minimally invasive incisions, which is beneficial in regard to infections (including mediastinitis) and bleeding from the incision and the operative field, leading to lesser usage of homologous blood.111 There is also improved cosmesis with these incisions, and postoperative pain seems to be considerably less than in patients with the median sternotomy. This can result in fewer requirements for pain medication and faster return to normal activity with less dependence on after-hospital stay and after-hospital care without compromising results.

Intracardiac Technique

Operation entails secure fixation of a valve prosthesis to the annulus by reliable suture techniques without damage to adjacent structures or myocardium and without tissue interference with valve function. Implantation should prevent injury to anatomical structures surrounding the mitral valve annulus. Figure 42-6 shows the proximity of important cardiac structures near the mitral valve annulus. These include the circumflex coronary artery within the atrioventricular (AV) groove, the left atrial appendage, the aortic valve in continuity with the anterior mitral curtain, and the AV node.

An accumulation of laboratory and clinical evidence indicates that preservation of papillary muscle–chordal attachments to the annulus is important for maintenance of left ventricular function. In patients with mitral stenosis with agglutinated, fibrotic chordae and papillary muscles, preservation of these structures probably has little effect on left ventricular dysfunction but does protect the AV groove from rupture by preserving the posterior leaflet. However, preservation of the posterior mitral leaflet may preclude use of an adequately sized prosthesis. If fibrotic, agglutinated chordae and the posterior leaflet are excised, placement of artificial Gore-Tex chordae to reattach the papillary muscles to the annulus may improve early and late preservation of cardiac output.112 In patients with mitral regurgitation, however, it is important to preserve as much of the papillary muscle and annular interaction as possible. This can be achieved by a variety of techniques, as shown in Fig. 42-7. The anterior leaflet may be partially excised and brought to the posterior leaflet113 (see Fig. 42-7A) or can be partially excised and “furled” to the anterior annulus by a running Prolene suture114,115 (see Fig. 42-7B).

Experimental and clinical evidence suggest that preservation of the conical shape of the ventricle is important to maintain normal cardiac output116–119 and that assumption of a globular shape from cutting papillary muscles is deleterious to left ventricular function. Furthermore, preservation of the posterior leaflet and chordae has reduced the incidence of perforation of the left ventricle and atrioventricular separation dramatically during mitral valve replacement.10,120,121

Suturing techniques vary according to the type of valve that is implanted. The bioprosthetic valve is inserted preferentially with the sutures placed from ventricle to atrium (noneverting or subannular). This has been shown to be the strongest type of suturing technique to the mitral annulus and is used with this valve and with the central-flow Starr-Edwards ball-and-cage valve (Fig. 42-8A).

To ensure adequate function of bileaflet or tilting-disk valves, everting sutures (atrium to ventricle to sewing ring) should be used (see Fig. 42-8B). This technique pushes the prosthetic valve out into the center of the orifice and minimizes any tissue interference of the prosthetic valve leaflets. This is particularly important if annular-chordal attachments are preserved. Teflon-pledgeted sutures, particularly with the thin sewing rings of the currently available bileaflet and tilting-disk valves, should be used. If a bioprosthetic valve is inserted, a dental mirror is used to ensure that no annular suture is wrapped around a stent strut. A running Prolene suture for implantation of mitral valves has been advocated by some surgeons.123,124 This technique makes a very clean suture line with minimal knots but runs the risk of valve dehiscence if an infection occurs.125

Before closure, the left atrial appendage is ligated by suture or stapled to prevent clot formation in patients with chronic atrial fibrillation, enlarged left atrium, or left atrial thrombus.126 The atrium is closed by a running Prolene suture, making sure that endocardial surfaces are approximated. If needed, a left atrial catheter can be inserted through the suture line.

Coronary bypass is the most common procedure performed with mitral valve replacement and should be performed first. This reduces lifting of the heart after the rigid mitral valve prosthesis is in place, which can cause rupture of the myocardium or the atrioventricular groove. This also allows cardioplegia to be delivered through the bypass grafts.

Tricuspid valve repair or replacements usually are performed after replacing the mitral valve. In these cases, the mitral valve often is approached through the right atrium and a transseptal incision. After the mitral valve prosthesis is in place, the septum is closed, and the aortic cross-clamp is removed before proceeding with the tricuspid valve procedure.127

When both the aortic and mitral valves are replaced at the same operation, most surgeons begin with excising the aortic valve before proceeding with the mitral valve procedure. When excising the anterior mitral valve leaflet, care must be taken not to injure the aortic annulus and the intra-annular region. The aortic valve then is sewn in after the mitral valve is in place.

Weaning off Cardiopulmonary Bypass

We use transesophageal echocardiography for every valve operation and particularly for mitral valves, where excellent images can be obtained. If transesophageal echocardiography is contraindicated (eg, because of esophageal disease), direct epicardial echocardiography can be used. The echocardiograms provide information about valve and left ventricular function, possible retained material in the left atrium including thrombus, and removal of intracardiac air.

A careful de-airing at the end of the operation is essential. The heart is vented through the left atrium and the ascending aorta and sometimes the left ventricle. Before the aortic cross-clamp is removed, the patient's head is lowered and the lungs inflated carefully to dislodge any air bubbles in the pulmonary vein. The operating table then is tilted from side to side, the left atrial appendage is inverted, and the cardiac chambers are aspirated if necessary. Once de-airing maneuvers are completed, and after the patient is completely rewarmed, venous return is partially occluded, and the heart is gradually volume loaded. Pulmonary artery pressures are monitored carefully. Pharmacologic agents, such as amrinone or dobutamine, particularly for right ventricular overload, are used frequently.

De-airing the heart after a minimally invasive approach is more complex because the heart is only partially visible through the 5- to 7-cm sternal incision, and the apex of the heart is not accessible. Flooding the operative field with continuous CO2 can be beneficial in reducing intracardial air.128 In addition, by filling the left ventricle with cardioplegia through the left atriotomy suture line, and manipulating the volume of the heart on bypass and alternating the position of the patient, air can be evacuated effectively.

Postoperative care is directed toward the resumption of normal cardiac output, respiratory function, temperature control, electrolyte management, adequate renal flow, and prophylaxis against bleeding. Patients with low cardiac output are managed with a variety of pharmacologic agents after providing adequate volume loading. Left atrial and especially pulmonary arterial catheters are particularly helpful in determining optimal balancing of volume loading and myocardial function in the first hours after operation.

Reduction in pulmonary interstitial fluid is pursued aggressively by diuresis in the intensive care unit in patients with severe pulmonary hypertension. Most patients with severe pulmonary hypertension can be extubated within 48 hours of surgery. Nutritional, respiratory, and general metabolic support is provided. Many patients with severe, longstanding mitral valve disease are cachectic and, despite preoperative nutritional support, are severely catabolic at the time of operation. These patients generally require longer periods of ventilatory support owing to lack of respiratory muscle strength. They need aggressive nutritional support with nasogastric hyperalimentation to increase respiratory muscle strength. In patients with severe pulmonary hypertension and cardiac cachexia who require prolonged intubation, tracheostomy may be necessary to reduce ventilatory dead space and facilitate faster weaning and better pulmonary toilet. Tracheostomy usually is performed by the end of the first postoperative week.

Postoperative atrial arrhythmias are so common that their absence is unusual. Arrhythmias vary from rapid supraventricular tachycardias, usually atrial fibrillation, to junctional rhythm and heart block. These arrhythmias are treated by pharmacologic agents, pacemakers, or both. If rapid atrial fibrillation cannot be controlled pharmacologically and is destabilizing hemodynamically, emergency cardioversion is done to improve cardiac output. Pharmacologic management of supraventricular tachycardia usually is required but may precipitate the need for a prophylactic transvenous pacemaker if severe slowing of the heart rate occurs.

Anticoagulation is prescribed for all patients undergoing mitral valve replacement with either a mechanical or bioprosthetic valve. In the first 6 weeks after operation, the incidence of atrial and other arrhythmias is high; thus, these fluctuating rhythms mandate anticoagulation even if the basic rhythm is sinus. In addition to rhythm concerns, the left atrial incisions and the possibility of stasis in the left atrial appendage justify full anticoagulation with warfarin for all patients. Some surgeons advocate immediate intravenous heparin until therapeutic warfarin doses can be reached.129,130 Low molecular weight heparin (LMWH) also can be used.131,132

The therapeutic international normalized ratio (INR) after mitral valve replacement is 2.5 to 3.5 depending on the type of valve, cardiac rhythm, and presence or absence of the aforementioned intraoperative risk factors for thromboembolism.2,126,130,131 Anticoagulation levels are in the low range for patients in sinus rhythm who received tissue valves. Patients who have mechanical valves need lifelong anticoagulation. Patients who have bioprosthetic valves are evaluated at 6 to 12 weeks for cardiac rhythm abnormalities. If they are in predominantly sinus rhythm, warfarin is stopped, and one aspirin tablet is given daily indefinitely. If the patient has continuous atrial fibrillation or fluctuating rhythms, anticoagulation with warfarin is continued. This is also true for patients with a history of previous embolism or in whom thrombus is found in the left atrium at operation.

Warfarin usually is started on the second postoperative day. Addition of aspirin, 80 to 150 mg daily, to the warfarin may reduce the risk of thromboembolism133,134 and may have a role in patients with prosthetic valves.135

Early Results

The hospital mortality for mitral valve replacement with and without coronary bypass grafting has decreased significantly since the inception of mitral valve surgery. The current risk (2006) of elective primary mitral valve replacement with and without coronary bypass grafting is 5 to 9% in most studies (range 3.3 to 13.1%).72 Operative (30-day) mortality is related to myocardial failure, multisystem organ failure, bleeding, respiratory failure in the chronically ill, debilitated individual, diabetes, infection, stroke, and, very rarely, technical problems.27,45 Mortality is correlated with preoperative functional class, age, and preexisting coronary artery disease.138

Published results on mitral valve surgery have improved in recent years,139 probably because of preservation of papillary muscles, preventing midventricular rupture,120,121 and preservation of the normal geometry of the left ventricle, which aids in the maintenance of early postoperative cardiac output.116,117,140 Mitral valve replacement and coronary artery bypass surgery 20 to 25 years ago had an associated mortality of about 10 to 20%.44,141 This mortality risk also has decreased as myocardial protection has improved with the use of blood cardioplegia and retrograde methods of administration.99,142 Some studies have indicated that the risk of combined mitral valve replacement–coronary artery bypass grafting is now no greater than that of mitral valve repair with an annuloplasty ring or mitral valve replacement without coronary artery bypass grafting.116,143 Other studies have shown significantly increased morbidity and mortality with the addition of coronary artery bypass grafting.48,144 Figures from the database of the Society of Thoracic Surgeons indicate that both reoperation and emergency operation increase operative mortality145 (Fig. 42-9).

Late Results

Functional Improvement

In more than 90% of patients after mitral valve replacement, functional class improves to at least class II. A small group of patients remains in class III or IV depending on left ventricular function before surgery or other coexisting morbidity.

Survival

The causes of late death in patients after mitral valve replacement are primarily chronic myocardial dysfunction, thromboemboli and stroke, endocarditis, anticoagulant-related hemorrhage, and coronary artery disease. The extent of left ventricular dysfunction and patient age, particularly if myocardial and coronary diseases are combined, also correlate with late mortality. The probability of survival after mitral valve replacement at 10 years is usually around 50 to 60% (range 42 to 81%)72 (Table 42-3). Long-term patient survival seems to be similar for patients with biologic and mechanical mitral valves.147–149Unlike patients with severe aortic regurgitation or aortic stenosis, arrhythmias seldom cause sudden death in patients after mitral valve replacement; however, a few patients die from thromboembolic stroke owing to chronic atrial fibrillation. The fact that more than 50% of patients after mitral valve replacement are in chronic atrial fibrillation increases the propensity for thromboembolic stroke despite anticoagulation and for mechanical valve thrombosis if the anticoagulation protocol is altered. In addition, patients with older types of prosthetic valves who receive higher-intensity anticoagulation may develop severe anticoagulant hemorrhage.150

In patients with bioprosthetic valves, one of the important determinants of mortality is reoperation secondary to structural valve degeneration53,54,72 (Table 42-4). Reoperative mitral valve replacement mortality has decreased significantly in the last 15 years to under 10%, even in patients who have required multiple mitral valve reoperations.142 At the Brigham and Women's Hospital, operative mortality was less than 6% for reoperative mitral valve operations from 1990 to 1995.142 Improved myocardial protection, earlier selection of patients for reoperation, and better perfusion techniques, including frequent femorofemoral bypass to protect the right ventricle during incision and dissection of the heart, are factors contributing to decreased mortality.116,142,152–154

Late Morbidity

The major morbidity in patients after mitral valve replacement is structural valve deterioration of a bioprosthetic valve and thromboembolism and anticoagulant hemorrhage with a mechanical prosthesis. Both valve types develop perivalvular leak and infection.

Thromboembolism

Thromboembolism is perhaps the most common complication of both biologic and mechanical mitral prostheses but is more frequent in patients with mechanical valves. Chronic atrial fibrillation and local atrial factors, discussed earlier, increase the risk of thromboembolism in patients with mitral prostheses. A number of recent studies have summarized the thromboembolic potential of various valves (Table 42-5), and it appears that the better the valve hemodynamics, the lower is the probability of thromboemboli. The incidence of thromboemboli in currently available bileaflet and tilting-disk valves is similar to that in bioprosthetic valves—about 1.5 to 2.0% per patient-year. Thromboembolism in patients with mitral valve replacement is lower in those with a small left atrium, sinus rhythm, and normal cardiac output. It is much higher in patients with a large left atrium, chronic atrial fibrillation, and the presence of intra-atrial clot.126,169 Thrombosis of a mechanical valve, once a feared complication of tilting-disk valves,170,171 is now relatively rare unless anticoagulation is stopped for any period of time. Valve thrombosis can be treated with thrombolytic agents if the patient is not in cardiogenic shock but requires surgery if the circulation is inadequate.172,173

Anticoagulant Hemorrhage

Bleeding related to anticoagulation is seen most commonly in the gastrointestinal, urogenital, and central nervous systems and usually is proportionate to the INR. The incidence of anticoagulant-related hemorrhage has decreased markedly with hemodynamic improvements in mitral valve prostheses. New valves do not require the intensity of anticoagulation of older prostheses. For example, the distinctive Starr-Edwards ball-and-cage valve requires an INR of 3.5 to 4.5.150 Patients with streamlined bileaflet or tilting-disk valves require an INR of between 2.5 and 3.5; thus, the incidence of anticoagulant hemorrhage is reduced significantly in the newer, hemodynamically improved prostheses.174Table 42-5 lists the incidence of anticoagulant hemorrhage with various bioprostheses and mechanical valves.

Structural Valve Degeneration

Structural valve degeneration (SVD) is the most important complication of the bioprosthetic valve. The probability of structural failure with currently available porcine valves (Hancock or Carpentier-Edwards) begins to increase 8 years after operation and reaches more than 60% at 15 years.51,53 This finite durability is a major impediment to long-term success of these biologic prostheses, even though the failure rate in a patient 70 years of age or older is significantly less than in younger age groups (Table 42-6). Structural valve degeneration presents as mitral regurgitation from leaflet tear, calcific mitral stenosis owing to calcification of valve leaflets, or both. The appearance of a new murmur with new congestive symptoms should prompt a noninvasive investigation of the prosthesis and elective re-replacement if dysfunction is documented. Structural valve degeneration leading to reoperation is the cause for at least two-thirds of the reoperations in patients with bioprostheses.152,153 The probabilities of structural valve degeneration at 5, 10, and 15 years of the six most commonly used biologic prostheses are shown in Table 42-2. With current quality controls, the incidence of structural valve degeneration is virtually zero for bileaflet, tilting-disk, and ball-and-cage valves.

Perivalvular Leak

Perivalvular leak is an uncommon complication that usually depends on technical factors. Patient-related factors such as endocarditis and calcifications involving the annulus are also important. Perivalvular leak usually causes refractory hemolytic anemia in contrast with the milder chronic hemolysis seen after implantation of some of the mechanical valves, especially the tilting-disk valves.175

Because of improved surgical techniques and the use of Teflon pledgets, the incidence of perivalvular leak has fallen and is about 0 to 1.5% per patient-year for both mechanical and biologic valves.158,168,176 Perivalvular leak is slightly more common with the bileaflet valve than with the porcine valve because of the need for the everting suture technique and less bulky sewing ring.177,178 Surgery should be offered to all symptomatic patients and even patients with mild symptoms who require blood transfusions.176

Endocarditis

Mitral valve endocarditis is considerably less common than aortic prosthetic valve endocarditis,179 but when it does appear, it may present as septicemia, malignant burrowing infections, abscess formation, and septic emboli. This often results in difficult management problems related to timing of operation, type of operation, ability to fix the prosthesis securely, and operative and late survival. With better antibiotic prophylaxis at the time of mitral surgery and improved prophylaxis for all patients having dental or other surgical procedures, the incidence of endocarditis has been reduced substantially.

The incidence of prosthetic endocarditis is highest during the initial 6 months after surgery and thereafter declines to a lower but persistent risk.1 The probability of freedom from this morbid event is shown in Table 42-7 for both mechanical and bioprosthetic valves. Biologic and mechanical valves have a similar incidence of endocarditis, except for the initial months after valve implantation, when mechanical prostheses carry a greater risk of infection.184

The diagnosis of prosthetic endocarditis is made by the presence of symptoms, appearance of a new murmur, a septic embolus, or large vegetation on echocardiogram. Blood cultures usually are positive, although a small percentage of patients have culture-negative endocarditis. Echocardiograms may show a rocking motion of the prosthesis and the presence of vegetations. The most frequent organisms are still Streptococcus and Staphylococcus; the latter is usually hospital acquired. Antibiotic therapy depends on the sensitivity of the organisms, but immediate high-dose intravenous therapy must begin as soon as possible. Experience indicates that a number of patients with bioprosthetic valvular endocarditis can be “cured” of low-potency organisms such as Streptococcus. However, it is unlikely that antibiotics alone can sterilize more virulent mitral valve infections, particularly Staphylococcus. These infections usually require urgent and sometimes emergent surgery because of invasion of the cardiac exoskeleton.

The indications for surgical intervention in mitral valve prosthetic endocarditis are persistent sepsis, organism, congestive failure, perivalvular leak, large vegetations, or systemic infected emboli.37,38,185 Operative technique is similar to other mitral procedures with respect to anesthesia, monitoring, cardioplegia, left atrial incision, and exposure of the valve. Critical to successful resolution is the complete excision of the prosthetic device and débridement of all infected tissues. Surgical technique is described in Chapter 43. Postoperative care should include at least 6 weeks of appropriate intravenous antibiotics. Hospital mortality is related primarily to ongoing sepsis, multisystem organ failure, or failure to eradicate the local infection and subsequent recurrent perivalvular leak.187,188 Recurrence of infection depends on the type of organism and the surgeon's ability to remove all areas of infection completely. Recurrence of infection is the single most important long-term complication.

Patient–Prosthesis Mismatch

The concept of patient–prosthesis mismatch is not new, but was first described in relationship to aortic valve replacement by Rahimtoola in 1978.189 Since his introduction, the deleterious effects of patient–prosthesis mismatch in LV remodeling, function, and early and late survival with aortic valve replacement have been well documented. The notion of mitral valve patient prosthesis mismatch (MVPPM) in the adult was first proposed in a case report in 1981 again by Rahimtoola.190 Most MVPPM was described in relation to pediatric mitral valve surgery, resulting in re-replacement rates approaching 30%, largely the result of somatic growth versus fixed prosthesis size. Recently, the possibility of such patient prosthesis mismatch in the mitral position has become an area of increased interest among adult cardiac surgeons. At this time, MCPPM has been studied through in vitro pulse generator analysis demonstrating that an index geometric orifice area (IGOA) less than 1.3 to 1.5 cm2/m2 could potentially result any patient having a high postoperative transprosthetic gradient. Similar studies performed in the clinical arena suggest that an index effective orifice area (IEOA) less than 1.3 to 1.5 cm2/m2 as measured by the continuity equation may increase the risk for MVPPM. Using this criterion, Lam and coworkers demonstrated a potential for MVPPM of roughly 32% in a recent series of 884 patients. In this population, patients at elevated risk for MVPPM measured at IEOA 1.0 to 1.25 cm2/m2 were at a significantly higher risk for development of postoperative congestive heart failure than those with IEOA greater than 1.25 cm2/m2. Although a weak association with CHF and pulmonary hypertension was noted, no direct correlation was found between MVPPM and the development of postoperative pulmonary hypertension.

Jamieson and coworkers recently evaluated the potential impact of MVPPM and long-term survival.191 In contrast to prior reports, this study of nearly 2500 patients refuted the notion that MVPPM had any relation to mortality. Instead predictors of overall mortality included age, New York Heart Association III or IV, competent coronary artery disease, ventricular dysfunction, prosthesis type, BMI, and pre-existing pulmonary hypertension.

Certainly some degree of patient prosthesis mismatch may occur as a result of mitral valve replacement. The surgeon should be aware of the potential for MVPPM, particularly in the setting of particularly small annulus size in patients resulting in the need for implantation of a small prosthetic device resulting in a high transprosthetic gradient. As such, foreknowledge of the devices specific IGOA may assist allowing the surgeon to choose a device that might minimize MVPPM. Unfortunately, unlike the aortic position, in which aortic root enlargement will allow for valve upsizing, the mitral position cannot be corrected as simply. As a result, in patients with particularly small mitral valve area, some degree of patient prosthesis mismatch may be unavoidable. This phenomenon remains an area of active clinical investigation. Its resolution will require larger meta-analysis.

Mitral valve replacement by mechanical or bioprosthetic valves revolutionized the care of patients with severe mitral valve disease. Reconstructive operations of the mitral valve now have assumed an equally important role for mitral regurgitation. A number of advanced lesions of the mitral valve still require mitral valve replacement with reliable devices. The bileaflet, tilting-disk, and ball-and-cage prosthetic valves are extremely reliable in terms of durability but require long-term anticoagulation and have a high risk of thromboembolism or thrombosis without anticoagulation. Bioprosthetic porcine valves, conversely, in patients in sinus rhythm do not need long-term anticoagulation and are used mainly in elderly patients who are not likely to outlive the valve and in women who plan to become pregnant and do not wish to accept the risks of warfarin or heparin. The long-term durability of these valves is limited, and the probability of valve failure at 15 years is at least 40%. Improvements in mechanical valve design and biologic valve preservation of collagen structure and resistance to calcification are ongoing and are the hope for the future. In addition, there is renewed interest in homograft mitral valves, which may offer better long-term durability, as has been observed with cryopreserved homograft aortic valves. Improved valve design and development of better biomaterials eventually will improve clinical results; however, current FDA restrictions on the development and evaluation of new prosthetic valves have an important impact on this process.