Long term outcomes primarily depend on left ventricular function
Primary mitral valve disease involves damage to leaflet or chordal tissue,1 whereas functional (or secondary) mitral regurgitation (MR) typically involves a combination of mitral annular dilatation and leaflet restriction caused by ventricular dysfunction in patients with normal leaflets and chordae. Assessing the severity of regurgitation in secondary MR is more difficult than in primary MR, as the regurgitant orifice area is often underestimated by echocardiography because of its crescent shape during systole. Left ventricular stroke volume is usually reduced in secondary MR, so that lesser degrees of regurgitant volume are more significant.
Mitral annular dilatation can follow left ventricular dilatation caused by dilated cardiomyopathy or myocardial infarction. Mitral and tricuspid dilatation can also occur as a consequence of atrial dilatation resulting from long-standing atrial fibrillation. This usually affects the tricuspid to a greater extent, but can still cause symptomatic MR that requires surgery.
Functional MR is thus a left ventricular condition,2 and anti-failure medical therapy is the critical first component of management. Some patients present with dyssynchrony of ventricular contraction, and can benefit from biventricular pacing.3 The traditional technique for repair is to perform a mitral annuloplasty, which aims to bring the posterior (free wall) of the annulus forward and thereby move the posterior mitral leaflet closer to the anterior leaflet, enlarging the coaptation zone. This works well on the operating table, but with time the left ventricle can re-dilate, which in many patients results in the posterior leaflet being drawn further down than can be compensated by the annuloplasty. In fact, regurgitation can be increased in some patients as tethering of the leaflets becomes more prominent, pulling the edges of the leaflets into the ventricle and decreasing the amount of coaptation between the leaflets.
Mitral valve replacement has historically been associated with a higher mortality rate than mitral valve repair across all categories.4–6 Data from the National Cardiac Surgical Database6 indicate a mortality rate in Australia of about 1% for isolated mitral repairs (largely myxomatous disease), but 4–6% for mitral valve replacement (myxomatous and rheumatic disease). Mortality increases markedly if concomitant coronary surgery is undertaken: coronary grafting with valve repair is associated with 4–6% mortality, or 8–12% if the valve is replaced.6 Some of this mortality is due to selection bias, but some is directly associated with the surgery. Placing a rigid, circular prosthetic valve in a dynamic, D-shaped muscular orifice can adversely affect ventricular function, as can failure to preserve papillary muscle continuity with the mitral annulus.
The mortality associated with surgery for functional MR depends on the degree of ventricular impairment, but ranges between 5% and 20% or more for patients with severe impairment. However, the results of valve repair in functional MR have been variable, with about one-third of patients experiencing the return of significant regurgitation within 3 months of surgery; on the other hand, the operative risk is lower than for valve replacement.4,5 The recurrence of regurgitation is explained by the inability to control the size and shape of the left ventricle in many patients. If the regurgitation is caused by annular dilatation with minimal leaflet restriction, annuloplasty is a good option with acceptable results. If there is significant tenting (the coaptation plane of the valve is more than 1 cm below the plane of the annulus), annuloplasty alone will not reliably achieve a competent valve, and valve replacement is preferred. Left ventricular size is the main factor, with an end diastolic diameter greater than 65 mm associated with poorer long term survival and repair results.
Other techniques for repairing the valve have been investigated. Some surgeons place a tape around the bases of the papillary muscles and draw them together to reverse the displacement caused by left ventricular dilatation.7,8 Techniques for internally splinting the left ventricle with Gore-Tex neochords or using external ventricular wraps have been tried. Other surgeons have attempted to enlarge the anterior mitral leaflet with a patch in order to move its coaptation edge back to meet the displaced posterior leaflet.9 A frequently used alternative is the edge-to-edge repair introduced by Alfieri, which can be used in both primary and secondary MR.10–12
There has been much debate about which approach achieves the best outcomes, a discussion complicated by the variations in results and case selection in published reports.
Patients with ischaemic cardiomyopathy can present for coronary surgery with any degree of MR, or with heart failure resulting from ventricular dysfunction and MR. Patients undergoing coronary artery surgery with moderate regurgitation can be managed with an annuloplasty during coronary artery surgery. A recent update on a large randomised trial13 has confirmed this consensus but questioned the benefit of the approach, the authors finding that survival and re-admission for heart failure were the same at 2 years, regardless of whether the valve had been repaired. Adding mitral valve repair to the coronary surgery prolonged the operation time, as expected, but also increased the stroke rate and atrial arrhythmias. Repair resulted in less MR, but coronary grafting alone often achieves this outcome. Ventricular re-modelling was better, however, in patients with less regurgitation. The problem is that it is difficult to define before the operation which ventricles will improve with revascularisation alone, and therefore which need mitral repair. It is notable that moderate to severe regurgitation was much more common if repair was not undertaken.13
Most importantly, long term outcomes depend primarily on left ventricular function. Patients with moderate to severe ischaemic MR who are having coronary surgery need a competent mitral valve. They usually have severe ventricular dysfunction and will not tolerate a failed repair, so that replacement has been the preferred option in most cases. The mortality risk for this subgroup is high, and mitral valve repair should be reserved for surgeons with a particular interest and experience in this field.14
Dilated cardiomyopathy can lead to both annular dilatation and papillary muscle displacement.2 Mitral valve repair in this group is challenging, again because of the inability to control the size of the ventricle over time, leading to recurrent regurgitation. These patients can present with improved ventricular function when they develop MR, as the ventricle empties into the pulmonary circulation. Left ventricular function deteriorates when the mitral valve is made competent, which increases the afterload against which the ventricle must eject its blood.
Severe MR is a common late feature of the dilating failing heart. Surgery in these patients is associated with a high risk of mortality, and should only be undertaken when supported by resources for managing heart failure after the operation, including short or long term ventricular support devices and, possibly, cardiac transplantation, depending on the clinical setting.
Cardiac surgery on patients with irreversibly damaged ventricles can be very risky, with long post-operative recovery periods. To reduce this morbidity, new technologies are being introduced. The Alfieri suture, for example, attaches the centre of the two mitral leaflets to limit abnormal leaflet motion, thereby improving mitral valve competency.10 A percutaneous catheter-based version of this approach is now used, the MitraClip; it is inserted in the catheter laboratory under fluoroscopic and echocardiographic guidance. It usually improves both regurgitation and functional class by one to two grades, which can be a major clinical improvement in this difficult to treat group of patients.15 However, many cases are not anatomically suitable for the MitraClip, and the patients need a replacement valve. Several new mitral implants currently undergoing trials are placed through the apex of the left ventricle, but do not require cardiopulmonary bypass, and can therefore reduce surgical stress for the patient. Time is needed to determine how effective they are with respect to ease of implantation and the relief of symptoms, as well as to ascertain any survival benefit.
Conclusion
Functional MR is usually caused by ventricular dysfunction. Mild to moderate degrees of MR can be readily treated with standard annuloplasty techniques. Their benefit, however, is questionable, and the operation should only be undertaken when performing coronary artery surgery if the additional operative risks can be justified. Moderate to severe regurgitation is more reliably corrected by replacing the valve, either in isolation or together with coronary surgery. The degree of left ventricular dysfunction determines the quality of the long term outcome, making these patients particularly challenging during the post-operative period. Newer catheter-based and off-pump techniques are being introduced that will help palliate the hazard for these high risk surgery candidates.