Is isolated aortic valve replacement sufficient to treat concomitant moderate functional mitral regurgitation? A propensity-matched analysis

Surgical aortic valve replacement (AVR) remains the most common valvular operation with over 50,000 procedures performed in the United States in 2013 alone [1]. In addition to correction of primary aortic valve pathology, current American Heart Association guidelines recommend mitral valve intervention for patients presenting with concomitant severe functional mitral regurgitation (fMR) as a result of their aortic valve disease [2‐4]. However, according to published literature, up to two-thirds of patients with aortic stenosis or insufficiency (AS, AI, respectively) can present with moderate or less fMR, and significant debate persists on whether these patients warrant mitral valve repair at the time of AVR [5‐12].

The traditional, conservative perspective assumes that non-severe fMR regresses following isolated AVR due to left ventricular reverse remodeling and removal of afterload obstruction, which subsequently leads to improved mitral leaflet coaptation. Although this phenomenon occurs to some extent following AVR, some reports suggest that patients may still be left with clinically significant fMR even at late follow-up, which may lead to worsened long-term survival [13‐15]. However, these studies are limited by small sample sizes and lack of extended follow-up, and fail to yield definitive conclusions. Therefore, the aim of this study is to determine the impact of uncorrected moderate fMR at the time of AVR on late survival in comparison to patients with no or mild fMR in a large patient population. These results may have important clinical implications in defining the appropriate treatment strategy for patients with combined aortic and moderate functional mitral valve disease, particularly in the modern era of transcatheter valve therapy.

Concomitant fMR in patients presenting for surgical AVR remains a challenging clinical problem. Although the predominant opinion is that fMR should improve following correction of aortic valve pathology, it is not clear that postoperative relief of left ventricular pressure-volume overload is sufficient to cause significant regression of moderate fMR, which may subsequently limit functional status and postoperative survival. While it seems that mild preoperative fMR will regress following isolated AVR, studies have shown that many patients with moderate preoperative fMR still demonstrate a clinically significant level of postoperative regurgitation on follow-up [6‐23]. Given the lack of well-defined treatment guidelines for these patients and small sample sizes used in prior studies, further investigation of the late effects of isolated AVR on moderate fMR is required.

The current study is, to the best of our knowledge, the largest single-center experience and the only propensity-matched analysis in this patient population. We have demonstrated that unaddressed moderate to moderate-severe (2–3+) fMR at the time of isolated AVR leads to equivalent perioperative survival but worsened medium- and long-term survival compared to patients with no or mild preoperative fMR. We did not detect a difference in postoperative complication rates or postoperative length of stay between groups. One-year echocardiographic follow-up revealed that patients with 2–3+ preoperative fMR had slightly but significantly worse residual MR compared to patients with mild or no preoperative fMR. While the post-AVR left ventricular reverse remodeling may improve fMR severity to some degree, our findings show that preoperative 2–3+ fMR in isolated AVR patients presages poorer late survival.

Review of our overall cohort analysis showed that patients with 2–3+ fMR are sicker than patients with no or mild fMR. Patients in Group MR were significantly older with worse renal and cardiac function, as evidenced by higher rates of congestive heart failure and lower ejection fractions. Group MR patients also had more severe TR, suggesting higher degrees of pulmonary hypertension and right ventricular dysfunction. Several prior studies addressing this population demonstrated similar findings [9, 11, 14]. Although it is possible that the worse prognosis of Group MR patients in the overall analysis is a direct effect of a greater degree of fMR, it is more likely that these patients simply have more advanced cardiovascular disease resulting in worsened overall survival. While our propensity-matched analysis allowed us to control for some of these baseline differences to more specifically evaluate the effect of preoperative fMR, it is worth noting that patients presenting with 2–3+ fMR in the setting of aortic valve disease generally have more advanced cardiac disease than their counterparts with no or mild fMR, and should be treated accordingly.

While many studies have demonstrated that severe MR should be repaired during concomitant AVR, controversy remains concerning the correct management of moderate and moderate-severe MR [2‐4, 19‐24]. Late survival in prior studies comparing patients with 2–3+ preoperative MR to patients with no or mild MR has varied. Given that the question at hand is whether or not isolated AVR is sufficient treatment for these patients, those studies that found a survival difference generally concluded that mitral intervention should be considered in patients with moderate or greater preoperative MR [6, 7, 13, 19‐23]. However, several other studies, including the largest single-center study from the Mayo Clinic (n = 190), found no difference in survival between the two groups, suggesting that mitral intervention may be unnecessary [8, 11]. Nonetheless, large baseline demographic differences among groups are present in these studies, fundamentally confounding interpretation of the data. Given the clinical equipoise in the conclusions of prior studies and the potential that moderate fMR may simply be an indicator of more advanced disease rather than a causative entity leading to poorer late survival, a propensity-matched analysis was essential to control for key baseline differences and remove confounding comorbidities. After propensity-matching, our data suggest that a lower bar for mitral intervention in these patients may be warranted. Further study into specific patient subgroups may be necessary to clarify the tradeoff between additional perioperative risk and late mortality with combined aortic and mitral surgery.

Our data suggest that patients with moderate, concomitant fMR undergoing AVR would benefit from mitral intervention, although the logical follow-up question to that observation is to ask whether mitral valve repair or replacement would confer the lowest perioperative mortality and the highest late survival in this population. A Cardiothoracic Surgery Trials Network study from Acker and colleagues evaluated the performance of mitral valve repair vs. replacement for patients with ischemic MR not undergoing AVR and found no difference in perioperative or 1-year survival, but the recurrence rate of moderate or severe MR was higher in the repair group [25]. Although not directly applicable to patients with combined aortic and mitral valve disease, follow-up of this randomized trial could help to predict which mitral intervention confers a greater survival benefit to patients.

Few studies exist that directly address AVR/mitral repair vs. AVR/mitral replacement in the fMR subgroup. In 2003, the Cleveland Clinic group published a study addressing all patients who underwent combined AVR with either mitral valve repair or replacement for any reason, not solely for fMR [26]. While there was no difference in perioperative survival between mitral repair and replacement, they found that concomitant mitral repair with AVR resulted in significantly improved survival at 5, 10, and 15-year follow-up compared to mitral valve replacement with AVR [26]. However, only 7% of the total population had fMR in the Cleveland Clinic series, which again highlights the need for a large randomized trial evaluating the appropriate therapy for patients with aortic valve disease and significant concomitant fMR [26].

There are several limitations to our study. It is a retrospective, single-center study that reflects the treatment biases of our clinical team. We have attempted to limit preoperative baseline differences by propensity matching, but it is possible that our propensity-matched groups do not accurately reflect the true population based on unmeasured covariates. Follow-up survival information did not include the cause of late death, so all comparisons of mortality reflect all-cause mortality and not death from cardiovascular causes. Finally, formal quantification of MR using more advanced echocardiographic parameters, such PISA derived metrics, will be necessary in future studies; these measurements are now a standard part of our echocardiographic analysis, but were not available for the majority of this patient dataset during our study period.

In conclusion, patients undergoing isolated AVR with concomitant moderate and moderate-severe preoperative fMR have worse medium- and long-term survival, and continue to have elevated MR severity at 1 year, compared to patients with no or mild preoperative fMR. These results indicate the need for mitral valve intervention at the time of AVR in patients with moderate or greater fMR. Given our results and those from prior studies, a randomized trial is needed to definitively clarify the optimal treatment strategy for this population, and to determine whether mitral valve repair or replacement at the time of AVR for moderate or greater preoperative fMR is superior.