Long-Term Outcome of Active Surveillance in Severe But Asymptomatic Primary Mitral Regurgitation

Early elective surgery in asymptomatic severe primary MR?

The optimal management strategy for patients with asymptomatic severe primary MR is controversial. Although some studies describe a potential advantage of early elective surgery (7,8,13), they need to be interpreted carefully (14).

Data from MIDA (Mitral Regurgitation International Database) suggested that early surgery was associated with a survival benefit in patients with flail leaflets when compared with “watchful waiting” (7). However, this was not a randomized study, having the limitations of a registry. It was based on an arbitrary definition of watchful waiting characterized by “medical management during the first 3 months of follow-up” and did not involve systematic structured follow-up in the reference centers. More importantly, in the “medical management” group, a “minimal subjective manifestation” (i.e., symptoms) was present in 32% of the patients, and 19% of the patients had a class II indication for surgery. In the registry reported by Kang et al. (8), early surgery was associated with lower event-rates and mortality but follow-up was not very strict because it included telephone interviews besides annual visits in the outpatient clinic. Interestingly, 17 patients in the early surgery group were lost to follow-up; 3 deaths that included 2 strokes and 1 infection were not taken into consideration because they were judged to be of “noncardiac” origin. However, events in the “conventional group” included heart failure as an indication for surgery in 22 patients and cardiac deaths in 6 patients who had already developed symptoms. In a more recent analysis with larger patient numbers, Kang et al. (15) reported no difference in overall mortality between early surgery and watchful waiting. Thus, the reported registries do not provide conclusive evidence in favor of early surgery.

Several additional caveats have to be considered in the context of early elective mitral valve surgery in asymptomatic patients.

Quantification of MR severity requires experience and may be challenging. Although the use of quantitative parameters such as the effective orifice area is recommended, it has been shown that interobserver agreement of quantitative parameters for the grading of MR is only close to 50% (16). Furthermore, even when using an integrative and quantitative approach, agreement between echo and cardiac MRI was only modest (10). Consequently, MR quantification in asymptomatic patients without signs of LV or LA dilatation may be overestimated. In a prospective study of patients with MR that were referred to mitral surgery, severe MR was confirmed using MRI in only 12 of 38 of patients who underwent surgery (9).

The outcomes of mitral valve surgery may be inhomogeneous. Experienced groups may achieve excellent repair rates >90% with a very low operative mortality <1% (17). However, in real world practice, up to 50% of the patients receive a mitral replacement, resulting in higher operative mortality (18) and exposure to prosthetic valve associated risks. Importantly, repair rates, operative mortality, and morbidity depend on hospital procedural volumes (19). Operative mortality may be <1% for high-volume centers and as high as 3% for low-volume centers (19). In addition, the combined risk of mortality and major morbidity is close to 10% even in high-volume centers (19). A high hospital volume alone was found not to guarantee successful mitral repair, denoting the “lottery of mitral valve repair surgery” (20). Ultimately, apart from hospital volume, individual surgeon-specific volume is essential; in an Society of Thoracic Surgeons database analysis of patients undergoing isolated mitral valve surgery, a strong correlation between surgeon-specific annual mitral valve volume with a probability of mitral repair was found with a predicted repair rate of 80% for 50 annual procedures (18). Strikingly, the median number of annual mitral valve operations performed per surgeon was only 5. In addition, the overall experience (and learning curve) of an individual surgeon needs to be considered, optimal outcomes being reliably achieved only after more than 300 procedures (21). Other aspects that need to be considered include the durability of repair with a rate of recurrence of at least moderate MR of 1.5% to 4% per year (22,23) and an annual reoperation rate of 0.5% to 2% (24,25). Thus, a more liberal approach to surgery may be valid for experienced surgeons in high-volume centers after careful evaluation. On the other hand, these considerations do not warrant a general recommendation of early elective surgery in asymptomatic patients with preserved LV function who are in sinus rhythm.

Excellent outcomes with active surveillance

Careful monitoring and follow-up until established indications for surgery, as defined in current guidelines are reached, is thus a valuable strategy. We have previously reported the safety of such a strategy, regarding perioperative and mid-term outcomes (6). The present study confirms that also the long-term outcomes of such an active surveillance strategy assessed in a large patient population are excellent. In comparison with the general population, no inferior long-term survival was found at 20 years and early survival was even better in the study population. This might be explained on the one hand by referral bias that potentially excludes patients with multiple comorbidities, especially those with advanced malignant disease, and the fact that vascular risk factors such as hypertension, hypercholesterolemia, and diabetes are promptly addressed in the course of the systematic patient workup in the HVC on the other hand. Furthermore, coronary artery disease was systematically screened and treated in patients who eventually developed an indication for surgery.

Importantly, the rates of asymptomatic LV dysfunction were low (7% of indications for surgery) and surgical outcomes were good, as expressed by a repair rate of 86%, a rate of residual MR comparing well with previous studies (26), a low surgical complication rate, and a low incidence of post-operative functional impairment. Post-operative LV dysfunction occurred in 15% of operated patients, a rate markedly lower than reported in previous series of patients who underwent surgical correction of MR (27).

In the setting of severe symptomatic MR (NYHA functional classes III and IV), sudden death is frequent with a reported rate of 7.8% per year (28). However, in the present study, only 2 cases of sudden death occurred in asymptomatic patients with preserved LV function, resulting in a sudden death rate of 0.1% per year, thus being comparable with the expected sudden death rate in the general population (29).

The concept of active surveillance in the setting of a HVC

In the present study, a strategy of watchful waiting was evaluated. It involved clinical and echocardiographic follow-up examinations performed in the setting of a HVC at a tertiary reference center every 6 months. Given the nature of this strategy, the term of active surveillance seems more appropriate. Although it is true that it may involve an initial conservative management phase, it is not a “conservative management” strategy because indications for surgery are strictly assessed and applied.

Providing a dedicated setting to deliver the required specialized care is thus essential for asymptomatic patients with MR. Active surveillance indeed involves a structured approach with scheduled follow-up visits in an experienced setting that also allows for further work-up, if required. Of note, one quarter of the patients had an indication for surgery based on echocardiographic or ECG findings, despite being asymptomatic, highlighting the importance of comprehensive follow-up.

Indeed, both the recent American College of Cardiology/American Heart Association (4) and the recent European Society of Cardiology/European Association for Cardio-Thoracic Surgery guidelines (5) on valvular heart disease now recognize the importance of heart valve centers of excellence. In this context, active surveillance of patients with MR in a dedicated HVC allows one to achieve an improved quality of care by optimal follow-up and timely referral to surgery.

Study limitations

Only patients with chronic severe primary MR were included in the study and, therefore, the results cannot be extrapolated to other forms of MR.

Surgical indications for asymptomatic primary MR have evolved mildly over the course of the study with the addition of surgical indications in selected low-risk asymptomatic patients in the more recent guideline updates. However, in the current analysis none of the patients were operated on for those latter reasons.

It might be viewed as a limitation that the assessment of the symptomatic status may be challenging and exercise testing, which was shown to be a predictor of outcome (30), was not performed in all patients. At the same time, the physician-based assessment of the functional status with the use of exercise testing in selected patients according to clinical judgment yielded excellent long-term survival.

Exercise echocardiography (31) was not routinely performed because its predictive value is unclear and neither the European nor the American guidelines base a surgical indication on that test.

The concept of LV global longitudinal strain in patients with primary MR has only been introduced long after the study initiation and this parameter was not part of the study protocol.

As mentioned previously, quantitation of MR severity may be challenging, however, in the present study, an integrated approach was used to quantify MR as recommended by current recommendations (11,12) and, furthermore, the majority of the patients had flail leaflets, which is highly characteristic of the presence of severe MR (11,12).

Mitral annular dysjunction has been described to occur in patients with mitral valve prolapse (32) and a recent article sought to address this issue with a three-dimensional echo study (33). In the present study we focused on the established analysis of mitral valve morphology only.

Serum biomarkers were not routinely assessed in this patient population. At the same time, no consistent threshold of serum N-terminal pro–B-type natriuretic peptide concentration has been defined to select patients for early surgery (4,5).