Thoralf M. Sundt III , MD

The question of how best to surgically approach ischemic MR has been hotly debated among opinion leaders in the cardiac surgical community controversial topic for more than 10 years. Based on the clear superiority of repair over replacement for degenerative disease, the pendulum has swung strongly in favor of repair over replacement for this entity despite recognition that the valve itself is structurally normal but distorted by ventricular pathology. Proponents of repair point to retrospective studies demonstrating markedly lower operative mortality with repair than replacement, even when groups are propensity matched. The counterpoint minority view has argued that subtle selection bias cannot be entirely eradicated from retrospective surgical studies rendering treatment groups truly comparable, that contemporary replacement is superior to historical results given techniques for preservation of the chordal apparatus, and that recurrent MR is common (up to 30% of patients) after repair. This study serves as an important contribution in guiding our care of these very challenging patients, and is especially notable as an accomplishment by this surgical network as it is so difficult to conduct a prospectively randomized surgical trial.

The principal finding of this study was that, in contrast to common perception and the findings of prior retrospective studies, perioperative mortality was not statistically significantly different between groups. This would tend to substantiate claims that differences reported in retrospective studies were influenced by unmeasured covariates — or in clinical parlance, ‘the foot of the bed’ test. This should provide encouragement to those who argue that ‘a good replacement is better than a bad repair.’ Equally intriguing is the observation that the clinically relevant endpoint of HF by NYHA classification was not different at 1 year between groups. If, as proponents of replacement would argue, regurgitation recurs to a clinically significant degree after repair, one would expect this to be reflected in greater degrees of congestive HF. This suggests that even if you have significant MR in that one-third of patients it does not impact their symptoms.

While this study does not answer all of our questions, it remains a very important contribution and establishes a platform for future research with longer follow-up. We will be most interested in the 3- and 5-year results with late survival, degree of recurrent MR and symptomatic status of these patients.

Timothy Gardner, MD

It is important to bear in mind that the patients being studied had ischemic MR only, not primary valvular mitral disease. This subset of ischemic MR is a more challenging subset, clinically and surgically. There has been a paucity of randomized controlled trials, especially in this group.

It is important to note that 11 of the intent-to-treat repair patients required replacement at the time of surgery and three more had reoperation in the first year.

There is a clear bias among cardiac surgeons and even some cardiologists that repair is preferable to replacement even for patients with severe ischemic MR. I think there has been perhaps some confusion over the effectiveness of repair for degenerative MR vs. ischemic MR. The present trial results clearly refute this assumption. Although there were no big differences in clinical outcomes, there were failed repairs and significantly more recurrences at 1 year in repair vs. replacement patients. I think we can conclude that chordal-sparing replacement of the mitral valve is at least equally effective for severe ischemic MR and may be preferable based on the significant incidence of MR recurrence.

This is the first initial report of the randomized trial in patients with severe ischemic MR. It has a companion study, a 300-patient randomized trial of moderate ischemic MR in which revascularization alone is being compared to revascularization and mitral valve repair. Those 1-year results are expected soon.