PHYSIOLOGY OF NORMAL VALVE FUNCTION
The function of the mitral valve is determined by atrial and ventricular volumes and pressure gradients, atrial and ventricular muscular contraction and relaxation, and the structural and functional integrity of the leaflets and chords. At the end of systole when the mitral valve is closed and the aortic valve is open, the ventricle begins to relax. The aortic valve quickly closes and the period of isovolumic relaxation begins. When the ventricular pressure falls below that of the atrium, the mitral valve opens and the rapid or early ventricular filling phase begins. More than two thirds of the eventual stroke volume enters the ventricle at this time (1). As the pressure gradient equalizes, the period of diastasis begins where there is little flow and the leaflets tend to move toward each other. With atrial contraction, the remainder of the stroke volume (generally less than 25% of the total stroke volume) is pushed into the ventricle to complete filling. After atrial contraction, the atrial pressure falls, this, along with the associated increase in ventricular pressure from systolic contraction, rapidly forces the valve into its closed position, entering the period of isovolumic ventricular contraction. Until the aortic valve opens, the back pressure on the mitral valve pushes it into the atrium. Continued contraction pulls the valve ring and valve towards the apex, reducing atrial pressure and promoting atrial filling. At the completion of ventricular systole the valve apparatus moves back towards the atrium and the process begins again. Each anatomical part of the valve contributes to this process, which is further detailed in Chapter 2.
REGURGITATION AS THE FINAL COMMON PATHWAY
Mitral regurgitation is the final common pathway of many disease processes. Damage to the valve leaflets or support structures leads to incompetence of the valve whether minor or major, which causes progression through ventricular dilatation, annular dilatation and impaired annular contraction, leaflet and chordal damage to increasing degrees of insufficiency, and progressive ventricular dysfunction. Chronic insufficiency leads to pulmonary hypertension (at first reversible then irreversible), associated right heart dilatation, tricuspid regurgitation, and eventually right heart failure. Whatever the etiology, most patients with severe mitral regurgitation have a contributing component of annular dilatation that must
be corrected as part of the repair. One scenario in which annular dilatation may not be a component of the pathophysiology is acute ischemia-induced mitral insufficiency, though chronic ischemic insufficiency is generally associated with annular dilatation. This absence of annular dilatation in the acute setting makes successful repair less likely than in the chronic situation in which dilatation has occurred. Other scenarios in which annular dilatation may not be significant include endocarditis or chordal rupture with acute regurgitation and rheumatic disease in which the fibrotic disease process prevents annular enlargement.
The Development of Mitral Surgery
The understanding of mitral valve structure, function, and pathophysiology has evolved significantly over the past thirty years. The first reports of repair using an open approach on cardiopulmonary bypass were published in 1957 (2,3) with the first description of repair of incompetence caused by a ruptured chordae in 1960 (4). These predated the first successful mitral valve replacement reported in 1961 (5). The first mitral valve replacements required removal of all parts of the mitral valve apparatus, including the leaflets and chordae, to prevent obstruction of the valve prosthesis. Though successful at treating the primary valve pathology, the first generation mechanical prostheses were very thrombogenic and required aggressive anticoagulation. Subsequent generations of mechanical valves required less anticoagulation but still incurred a significant incidence of valve-related hemorrhage and thromboembolism. Concurrently bioprosthetic valves were being developed but long-term results in the mitral position in younger patients were disappointing, with high rates of reoperation for structural valve deterioration. Surgeons also noted that the ventricles became more spherical after removing the subvalvular apparatus and removal of the posterior leaflets and chordal structures was associated with postoperative rupture through the posterior ventricular wall. A spherical-shaped ventricle was less mechanically efficient than the normal bullet-shaped ventricle. It became evident that postoperative function was better with chordal sparing. With the development of prostheses most surgeons switched from repair to replacement. However, a few continued to develop techniques of repair. The earliest of these was a simple suture annuloplasty to reduce mitral circumference (6,7,8). Subsequently techniques for leaflet augmentation, repair of posterior chordae, and prosthetic ring support emerged (9,10).
The Benefits of Repair
With the availability of reliable, increasingly durable mechanical bioprostheses why pursue mitral repair? What is the justification for a procedure that might take longer, has a low but definite incidence of early failure, and requires the use of a collection of techniques to mold the valve to a successful correction?
Valve replacement is simpler, and across a population of practitioners, more consistently reliable. Mitral repair requires practice and experience to achieve consistently good results. Nevertheless, repair over forty years of progressive refinements has become amenable to more standardized proven approaches and proven to have comparable durability to replacement with a lower incidence of thromboembolism, infection, and hemorrhagic complications.
Indications for Operation
With the advent of mitral repair and its associated lower incidence of long-term complications and good long-term success, the indications for referral for mitral surgery have changed. Recent AHA/ACC guidelines suggest early intervention if repair is likely (Table 1.1) (11). Included in these recommendations is the variable of prediction of the likelihood of repair based on the surgeon's skill and experience, the disease process causing regurgitation, and the involved portions of the mitral valve. Specifically posterior leaflet prolapse due to degenerative (nonrheumatic) mitral valve disease or a ruptured chord, which is likely repairable, in contrast to involvement of the anterior leaflet, which requires a more complex repair. They emphasize that “the skill and experience of the surgeon are probably
the most important determinants of whether the operation is eventually performed.” Further elaborating that in general, “rheumatic and ischemic involvement of the mitral valve and calcification of the mitral valve leaflets or annulus diminish the likelihood of repair even in experienced hands.”
Given these qualifications, surgery is recommended in the following scenarios: severe regurgitation with congestive heart failure symptoms and normal left ventricular function (ejection fraction > 0.60 and end-systolic dimension < 45 cm) and asymptomatic or symptomatic patients with evidence of reduced left ventricular function, specifically an ejection fraction ≤ 0.60 and/or left ventricular end-systolic dimension ≥ 45 mm, emphasizing that ejection fraction should not be allowed to fall into the low normal range, noting that there is a strong correlation of reduced survival after surgery with lower preoperative ejection fraction. If mitral repair can be accomplished in patients with a low ejection fraction, deterioration of left ventricular function can be slowed or arrested with accompanying symptomatic improvement. Shin et al. demonstrated that though ejection fraction is reduced after successful repair, stroke volume is preserved, independent of preoperative ejection fraction (12).
Surgery for mitral regurgitation is more controversial in the absence of symptoms. No data is available to recommend this, though the recent onset of atrial fibrillation is considered a relative indication, especially if there is a high likelihood of repair. The presence of atrial fibrillation for more than 1 year and/or left atrial size greater than 50 mm is a
predictor of persistent atrial fibrillation after valve surgery (13). The addition of ablative operations to cure atrial fibrillation to the surgeon's armamentarium will impact this decision process, perhaps prompting earlier surgery for the combination of atrial fibrillation and mitral regurgitation.
TABLE 1.1 PERCENTAGE OF PATIENTS UNDERGOING MITRAL VALVE SURGERY WHOSE VALVES ARE REPAIRED
Recently David et al. reported late outcomes of repair for floppy valves to bolster these guidelines (14). They reviewed results in 488 consecutive patients, 199 with no or minimal symptoms and 289 with symptoms at the time of surgery. With an overall survival of 61%, the survival in asymptomatic patients was 76%, identical to the age-matched general population, whereas survival for symptomatic patients was 53%, less than predicted for the general population. At 15 years, freedom from reoperation was 91%, and freedom from significant regurgitation (greater than 2+) was 85% for all, 96% for asymptomatic, and 76% for symptomatic patients. However if reoperation is the primary indicator used to assess durability, repair failure may be underreported. Flameng et al. evaluated recurrence of regurgitation as a marker instead of reoperation (15). At 8 years survival was 90.9 ± 3.2% and freedom from reoperation 94.2 ± 2.3%. Rates of freedom from recurrence of nontrivial (> 1/4) and severe (> 2/4) mitral regurgitation are detailed in Table 1.2.
They calculated a linearized recurrence rate of nontrivial and severe MR of 8.3 and 3.7% per year. Excluding patients with risk factors for recurrence (chordal shortening and no annuloplasty ring or sliding plasty) the rates were 6.9 and 2.5% per year. Overall freedom from bleeding and thromboembolic events was 90.4 ± 2.7% at 8 years. They caution that durability for successful repair is not constant and should be taken into account when making surgical decisions.
When to Repair, When to Replace
Repair, in most scenarios, represents the best alternative. However, repair can be time-consuming leading to prolonged ischemic and cardiopulmonary bypass times. The decision to proceed with repair needs to be based on the likelihood of success and the long-term benefit of repair. In scenarios where the marginal benefits of repair are approximated by valve replacement, replacement may be preferable. For example, chordal sparing valve replacement provides similar mechanical advantages to repair (16) and long-term anticoagulation is not essential with a bioprosthetic valve. Replacement might be preferable in elderly patients who have no issues with durability of the prosthesis, or during emergency surgery where short ischemic time is important or repair is difficult and potentially unreliable. For example, though techniques are described to repair complete papillary muscle rupture, attempts at repair in the setting of an acute infarction are probably unwise (17). Similarly, the long-term results of repair of rheumatic valves and valves damaged by radiation (18) are significantly worse than repair for degenerative etiologies, thus rheumatic etiology might reduce the impetus for a complex repair. Durability of repair of degenerative valves can also be variable (15).
There is some objective evidence that valve repair conveys a survival benefit. Enriquez-Sarano et al. (19) compared repair to replacement for isolated mitral valve disease noting that multivariate analysis showed an independent beneficial effect of valve repair on overall and late survival, operative mortality, and postoperative ejection fraction. However, this survival
benefit was not demonstrable in octogenarians (20). Gillinov et al. (21) recently compared repair to replacement for degenerative mitral disease associated with ischemic heart disease, a category perhaps often considered with ischemic mitral disease. This is a situation where many surgeons would opt for primary valve replacement, particularly in the elderly with whom a bioprosthesis can be used. Using multivariate analysis and multiphase, hazard-function analysis they determined that after 2 years repair imparted a survival advantage. Therefore every effort should be made to perform repair, even ifbileaflet reconstruction is necessary. In contrast, Thourani et al. performed a retrospective case matched study in 1,250 patients, 625 mitral valve repairs matched with 625 mitral valve replacements. Repair was associated with reduced length of stay, improved in-hospital mortality, and 10-year survival, however repair did not improve 10-year survival in patients over 60 years of age or those requiring associated coronary bypass (22). This finding is at variance with other reports (23). Though results of repair are superior to replacement in all types of regurgitation not secondary to ischemic changes, the results of repair in ischemia-induced mitral regurgitation in patients over 60 are not clearly better than replacement (23). Moss et al. used propensity matching to eliminate selection bias in comparing repair and chordal sparing replacement, noting that over a median follow-up period of 3.4 years survival was better with repair but there was a trend to increased reoperation (24).
TABLE 1.2 RATES OF FREEDOM FROM RECURRENCE OF NONTRIVIAL AND SEVERE MITRAL REGURGITATION
Perhaps the definitive analysis of the long-term results of mitral valve repair were published by Carpentier in 2001; the first 162 consecutive patients treated between 1970 and 1984 for mitral insufficiency secondary to nonrheumatic disease (25). The 20-year survival was similar to that for the general population, 3 patients required early reoperation and 7 required late reoperation. All 65 living patients were NYHA functional class I or II.
Interest in repair of the mitral valve has evolved over time. This is based on the problems associated with prosthetic valves including hemorrhage and thromboembolism, endocarditis, pannus formation, and structural failure. A survey of trends in valve repair from the Society of Thoracic Surgeons National Cardiac Database (NCD) demonstrated a progressive increase in the percentage of valves repaired through the 1990s (Fig. 1.1) (26).
Despite this increase, comparison with a “gold standard” suggests that repair is not being performed as often as possible. Mohty and colleagues (27) from the Mayo Clinic compared 679 repairs and 238 replacements performed between 1980 and 1995. Recognizing a referral bias (patients with a potentially repairable valve might be preferentially referred) this experience suggests a standard for rates of utilization of repair. The overall rate of repair (74%) for isolated mitral insufficiency (1980 and 1995) is almost twice that of the national rate of 42.4% in the period from 1999 and 2000 (26). Of these, 55% were repaired by annuloplasty alone with no other valvular reconstruction (26). This clearly shows that valve repair has not penetrated as much as it should.
Figure 1.1 Trends in the use of mitral valve repair through the 1990s. (From Savage EB, Ferguson Jr. TB, DiSesa VJ. Use of mitral valve repair: analysis of contemporary United States experience reported to the Society of Thoracic Surgeons National Cardiac Database. Ann Thorac Surg. 2003; 75:820-825.)