Outcomes of patients with mitral annular disjunction undergoing mitral valve repair^†

Abstract

OBJECTIVES

Mitral annular disjunction (MAD) is an abnormal displacement of the posterior mitral leaflet into the left atrial wall, potentially leading to left ventricular dysfunction, malignant ventricular arrhythmias (VA) and sudden cardiac death. This study investigates the outcomes of patients with and without MAD undergoing mitral valve repair for valve prolapse (MVP).

METHODS

The study retrospectively collected a single-center experience from 2021 to 2023 on 326 consecutive patients undergoing mitral valve repair for MVP. Patients were divided into two groups according to the presence of MAD. After propensity score matching 1:1, two comparable groups of 50 patients were obtained. Primary endpoints included hospital survival and early failure of the repair. Composite secondary endpoint included major adverse cardiac events (MACEs) such as reoperation, residual regurgitation ≥2, severe postoperative left ventricle (LV) dysfunction requiring prolonged (>3 days) inotropic support, cardiac arrhythmias and overall survival.

RESULTS

After matching, there were no significant differences between the groups in terms of preoperative characteristics. Hospital mortality was 0% in both groups, and there were no significant differences in terms of early reoperation (0%) or residual mitral regurgitation ≥2 or major atrial/VA. Nevertheless, patients with MAD presented a greater need for prolonged inotropic and mechanical circulatory support (IABP/ECMO): No-MAD 0% vs MAD 10% (P = 0.050). However, the composite outcome at midterm follow-up was similar between the groups.

CONCLUSIONS

Mitral valve repair in patients with MAD was associated with a significantly higher incidence of early LV dysfunction requiring mechanical support. However, no difference was found in terms of survival at follow-up.

Open in new tabDownload slide

INTRODUCTION

Mitral annular disjunction (MAD) is a structural abnormality of the mitral valve annulus, characterized by a systolic separation between the ventricular myocardium and the mitral annulus supporting the posterior mitral leaflet, which is implanted directly on the left atrial wall. The absence of basal myocardial attachment to the annulus results in paradoxical bulging of the ventricular myocardium and the annulus, called ‘curling’, which forms the apical margin of the MAD trench. The abnormal systolic movement of the mitral annulus leads to mechanical instability of the mitral valve apparatus, with an increased stress tension on posterior papillary muscle and a detrimental effect on left ventricle (LV) systolic function [1].

MAD has recently gained attention because of its association with various cardiac diseases, including major ventricular arrhythmias (VA) and sudden cardiac death (SCD). The prevalence of MAD in the general population ranges from 7.2% to 8.7% [2]. However, depending on the diagnostic criteria and imaging techniques used, the incidence of MAD is notably higher in patients with mitral valve prolapse (MVP), ranging from 25% to 40%, particularly in patients with myxomatous valves [2].

In addition, MAD seems to be associated with the development of both atrial and VA [3–5], probably due to an anatomical substrate (fibrosis), which may cause heterogeneity in refractory period and conduction velocity of the myocardium. The correlation between malignant VA and MAD was first demonstrated by Basso et al. [6]: the incidence of arrhythmias in patients with MAD and MVP is remarkable, with VA observed in up to 34% of cases.

Arrhythmic mitral valve prolapse (AMVP) is a specific subgroup of MVP characterized by the presence of complex VA, particularly in patients with significant mitral regurgitation (MR) and those with underlying MAD [7]. AMVP is associated with an increased risk of SCD, emphasizing the clinical importance of understanding the interaction between MAD, MVP and arrhythmias. In this scenario, the role of surgical mitral repair, including annuloplasty, has not been investigated yet, but surgical annular stabilization may mechanically reduce abnormal annular motion, with a decrease in abnormal myocardial stress and atrial and VA. This study aims to investigate outcomes of patients with MAD and MVP undergoing mitral valve repair.

PATIENTS AND METHODS

Ethical statement

The study adhered to the ethical principles outlined in the Declaration of Helsinki. Written informed consent was obtained from all patients prior to their inclusion in the study. This study was approved by the Institutional Review Board of University of Brescia under approval ID 1815, dated 15 October 2022.

This was a retrospective, observational, single-center study aimed at evaluating the outcomes of patients undergoing mitral valve repair, with a specific focus on those with MAD. MAD was defined as a separation greater than 8.5 mm between the posterior mitral annulus and the left ventricular wall at the atrioventricular junction, measured during transoesophageal echocardiography (TEE).

Patients

The study enrolled 326 consecutive patients who underwent surgical mitral valve repair with annuloplasty by means of rigid and semi-rigid rings at University Hospital, Brescia, Italy, between January 2021 and December 2023. The study population was identified through the institutional database and included patients with mitral valve regurgitation with leaflet prolapse (MVP), with or without concomitant MAD, requiring surgical repair. The presence of MAD was detected by preoperative and intraoperative TEE. MAD was confirmed in 50 patients, while the remaining 276 patients were classified into the No-MAD group.

Patients younger than 18 years of age, those with mitral valve endocarditis or mixed mitral disease such as rheumatic/inflammatory were excluded. Concomitant procedures were not considered as exclusion criteria.

Preoperative assessment is described in the Supplementary Materials S1.

Endpoints and follow-up

Primary endpoints were in-hospital mortality and early failure of repair (defined as regurgitation ≥2) and overall survival at follow-up. Secondary composite endpoint included major adverse cardiac events (MACEs) defined as: reoperation, residual regurgitation grade ≥2, severe postoperative LV dysfunction requiring prolonged (>3 days) inotropic support, cardiac arrythmias and overall survival.

Follow-up was conducted up to March 2024 (median follow-up: 1.83 [IQR: 0.81–2.64] years) by phone interview or clinical examination.

Statistical analysis

Continuous variables were compared using independent Student’s t-test if normally distributed. For non-normally distributed variables, the Mann–Whitney U test was used. Categorical variables were compared using the chi-squared test or Fisher’s exact test as needed. To ensure comparability between groups, propensity score matching (PSM) 1:1 was applied using a nearest-neighbour matching method without replacement. Balance of the variables was checked by standardized mean difference, and a threshold of 0.1 was set as good balance. Matching was based on key pre- and intraoperative characteristics to control for potential confounders (Fig. 1). For propensity-matched data, continuous variables were compared using paired Student’s t-test if normally distributed, while for non-normally distributed variables, the Wilcoxon signed-rank test was used. Survival differences between the two groups were represented and compared using the Kaplan–Meier method with stratified log-rank test for propensity-matched patients. No missing values were present, as all preoperative characteristics were required for propensity matching, and postoperative outcomes were readily available from the patients’ charts.

A P-value ≤0.05 was considered statistically significant. Microsoft Office Excel (Microsoft, Redmond, WA, USA) was used for data extraction, and all analyses were performed in R, version 4.3.1 (R Software for Statistical Computing, Vienna, Austria).

RESULTS

From the initial cohort of 326 patients, two balanced groups of 50 patients each were obtained by PSM. Preoperative clinical and demographic characteristics of the study cohort after PSM are shown in Table 1: the two groups had comparable age (No-MAD: 58 vs MAD: 56 years, P = 0.575), BMI (No-MAD: 22.8 vs MAD: 22.9 kg/m², P = 0.407) and a median EuroSCORE II (P = 0.952). Patients’ comorbidities were balanced after PSM. Median ejection fraction was 63% vs 62% and median left ventricular global longitudinal strain (GLS) was −16% vs −18% for No-MAD and MAD, respectively (Table 2).

Intraoperative outcomes are shown in Table 3. No differences were found on the type of intervention (concomitant procedures), aortic cross-clamp time or cardiopulmonary bypass (CPB) time, nor in terms of surgery timing (elective/urgency). Conversely, minimally invasive right thoracotomy was adopted more frequently in the No-MAD (21 patients vs 33 patients, P = 0.027). The aetiology of all included patients was primary MR. Mitral valve repair in the MAD group required the use of a larger mitral ring (median values: MAD: 36 mm vs No-MAD: 32 mm, P = 0.003).

Table 4 presents the early outcomes. No in-hospital mortality was reported in either group. However, the use of temporary mechanical circulatory support (MCS) was significantly higher in the MAD group, with 5 patients (10%) in the MAD group and none in the No-MAD group (P = 0.050). All patients received intra-aortic balloon pump (IABP) while only one patient subsequently required an upgrade to extracorporeal membrane oxygenation (ECMO) due to worsening of the clinical condition. There were no significant differences between the groups in terms of blood transfusion rates (P = 0.999), surgical revision for bleeding (P = 0.674), stroke (P = 0.999), postoperative dialysis (P = 0.999). One patient in the MAD group had residual MR grade >2, which was not present in the early postoperative course.

Postoperative atrial fibrillation occurred in 24.0% (n = 12) of patients without MAD compared with 34.0% (n = 17) of patients with MAD, P = 0.378. Similarly, postoperative VA requiring pharmacological treatment were observed in 6.0% (n = 3) of MAD patients compared with none in the No-MAD group (P = 0.241). No patients required new pacemaker implantation postoperatively. There was a trend towards a longer hospital stay for MAD when compared with No-MAD (7 [6–8] vs 7 [6–9] days, P = 0.084). At a median of 1.83 years follow-up, survival rates were comparable between the two groups (log-rank = 0.317) with only one death in the No-MAD group due to the rupture of a pre-existing abdominal aneurysm (Fig. 2). The 2-year MACEs incidence was 4.2% ± 2.9% and 2.0% ± 2.0% in the No-MAD and MAD groups, respectively (log-rank = 0.529, Fig. 3).

DISCUSSION

In this study, we compared outcomes of patients with and without MAD who underwent mitral valve repair for MVP. The key findings of this study can be summarized as follows:

• After PSM, patients with MAD required a significantly larger median annuloplasty ring size for mitral repair compared to patients without.

• Patients with MAD had a greater need for prolonged inotropic and MCS (ECMO/IABP) in the postoperative period due to temporary heart failure.

• No in-hospital mortality was reported in either group.

• The presence of MAD did not jeopardize the possibility of a successful mitral repair.

• Short-term survival and incidence of MACEs were comparable in both groups.

MAD is a common finding in patients with MVP undergoing mitral valve repair, with a reported prevalence ranging from 16% to 31% [8–10]. This finding is consistent with our results (MAD prevalence 15.3%). It has been proposed that in patients with MAD, abnormal mitral annular dynamics leads to an increased tension and stress on the myocardial wall, in particular on the posterior papillary muscle. This condition may result in a localized myocardial fibrosis, which may be a key factor for the development of cardiac arrhythmias [11].

The presence of MAD introduces additional mechanical stress on the mitral valve annulus, contributing to its dilatation. In fact, MAD is defined by a separation between the mitral annulus and the ventricular myocardium, leading to abnormal systolic motion, particularly in the posterolateral region, with paradoxical end-systolic dilatation of the annulus. The ‘curling’ of the unsupported annulus further amplifies this stress, contributing to its progressive dilatation. As it stretches, the annulus loses its normal saddle shape, which impairs valve function and can lead to MR. This additional stress accelerates annular dilatation, with patients with both MAD and MVP reported to have a higher regurgitant fraction than those without MAD [12]. The finding of a more dilated mitral annulus in patients with MAD is corroborated by a study by Shechter et al. [13], who reported the outcomes of patients undergoing transcatheter mitral valve repair (TEER). In that study, patients with MAD had more extensive MVP, larger valve size and required more devices during the procedure to achieve a successful repair. Similarly, Gray et al. [14] showed that patients with MAD had more complex MVP and larger annular diameters compared to those without MAD. This was further supported by Essayagh et al. [5], who found that MAD in MVP patients significantly altered annular dynamics, resulting in significant expansion and excessive annular enlargement during systole, which affects leaflet coaptation. The impact of MAD on left ventricular function remains uncertain, particularly as it complicates the accurate estimation of left ventricular ejection fraction (LVEF). During systole, MAD is often associated with excessive thickening of the basal region, particularly in the posterior wall, as the myocardium bulges inward without its usual annular support. This abnormal motion can simulate hyperdynamic LV function, leading to an overestimation of contractility (LVEF) [15–17]. In our study, although LVEF was comparable between patients with and without MAD after mitral repair surgery (matched population), patients with MAD had a significantly higher need for mechanical support (ECMO/IABP) due to transient postoperative left ventricular dysfunction. All patients fully recovered and were weaned off support, with no mortality in both groups. Unlike Özyıldırım et al. [16], we found no significant differences in GLS between groups. However, the higher postoperative need for MCS may indicate that changes in the LV and annular dynamics after mitral valve repair with annuloplasty in patients with MAD may reveal an underlying ventricular dysfunction, which may manifest after surgery and may contribute to the onset of postoperative afterload mismatch syndrome. This may explain the high incidence of MCS in the study MAD group. Left ventricular dilation has been observed in patients with MAD and MVP, even when significant MR was absent [18, 19]. Potential mechanisms for this finding include an underlying connective tissue disorder such as Marfan syndrome, MR underdiagnosis, frequent unnoticed ectopic beats or an independent myocardial condition [20]. Similarly, also left ventricular dilation and globally reduced postcontrast T1 times, indicative of diffuse fibrosis, have been described [21].

Despite previous findings, the presence of MAD did not negatively affect 30-day mortality or valve reparability. No 30-day mortality was reported in this study. This is consistent with previous studies by Gray and Eriksson on patients with MAD [14, 22] and with the MVP repair registries and trials [23, 24]. We also observed excellent repair rates in both groups, despite several studies indicating that patients with MAD often have more complex mitral valve disease, with extensive leaflet scallop involvement, more frequent bileaflet prolapse, larger annular dimensions and higher degrees of regurgitant fraction [5]. Furthermore, Eriksson et al. demonstrated that greater degrees of disjunction are associated with more extensive leaflet segment involvement [22]. This finding suggests that in patients with MAD, surgical technique for annuloplasty may be tailored according to the anatomy of the MAD. As far as the presence of MAD is concerned, there is a lack of data comparing resection versus respect strategies with artificial chordae. Drescher et al. reported no significant differences in terms of repair outcomes, nor in terms of mitral prosthetic ring implanted [25]. Finally, several studies have reported that after mitral ring implantation, MAD is no longer visualized by echocardiography due to the complete collapse or the significant reduction of the disjunction area [14, 26, 27].

This finding has important implications for patients’ treatment choices. MAD has been associated with major VA, such as ventricular fibrillation or tachycardia, as well as SCD. The risk of VA and SCD significantly increases when the disjunction exceeds 8.5 mm [28]. The suggested mechanism underlying these arrhythmias may involve a valvular trigger acting on a pathological myocardial substrate, such as hypertrophy or fibrosis [6]. The curling motion of the postero-basal segments caused by MAD may create abnormal stress on the local myocardium [5, 6]. In response to this altered loading condition, focal myocardial hypertrophy and extracellular matrix deposition occur, resulting in local fibrosis, particularly near the papillary muscles [6, 29], which serves as an anatomical substrate for re-entrant VA [25].

In addition, the distal Purkinje fibres that run through the papillary muscles are susceptible to abnormal automaticity and early depolarizations. These triggered premature ventricular contraction (PVC), in association with an electro-anatomical abnormal substrate (fibrosis), may lead to electrical re-entry circuits [30].

Our study corroborates the hypothesis that following surgical mitral valve repair with annuloplasty by means of rigid and semi-rigid rings, abnormal papillary muscle traction and systolic mitral annular displacement are mostly corrected. However, in the perioperative period, 3 patients in the MAD group (6%) experienced VA, while no arrhythmias were observed in patients without MAD. This may be explained by the fact that anatomical correction of MAD cannot interrupt the pathological arrhythmic substrate in the early postoperative period, and the possibility to obtain a reverse remodelling of the abnormal myocardial substrate has not been demonstrated yet.

The previous finding aligns with the data from Gray et al., who also reported a higher incidence of VA in MAD patients postoperatively [14].

Nonetheless, at the 4-year follow-up after mitral valve repair, Essayagh et al. [5] found that the incidence of VA in patients with MAD was comparable to those without MAD, as were long-term survival rates. Conversely, in patients undergoing TEER, MAD was a significant predictor of mortality during follow-up, despite good perioperative outcomes in terms of repair [13]. This may suggest that repairing the valve without addressing annular stabilization may not treat abnormal mitral systolic motion, being ineffective in eliminating the pathological consequences of MAD.

Limitations

The single-center design as well as the retrospective nature may also represent a limitation of this study. The small sample size after PSM, mostly related to the small number of patients in the pre-match group, combined with the limited number of events, may reduce statistical power of the study and preclude a generalization of the study’s findings. Future studies with larger prospective cohorts are needed to confirm these findings. Additionally, the median follow-up period of 1.83 years may be insufficient to capture long-term outcomes or recurrences of MR and/or VA.

CONCLUSION

In our study, the presence of MAD did not compromise the success of mitral repair. However, MAD patients had a higher need for temporary MCS in the postoperative period (IABP and ECMO). Nonetheless, both early and late all-cause mortality appeared comparable in MAD and No-MAD patients, with no significant incidence of VA and SCD. In patients with MAD requiring mitral valve repair for MVP, restoring correct anatomical relationships is crucial to prevent increased stress on the chordal apparatus, which is a key factor in the development of papillary muscle fibrosis and VA. Using a rigid or semi-rigid annuloplasty ring is essential to eliminate the paradoxical motion of the posterior mitral annulus by anchoring it to the fibrous trigones, the most stable part of the mitral annulus. Achieving a mitral repair without residual regurgitation, while restoring normal valve dynamics and stabilizing the MAD, provides a durable outcome by ensuring valve competence and disrupting possible pro-arrhythmic mechanisms. Mastery of all mitral repair techniques, including tissue resection as appropriate, is fundamental for optimal results.

SUPPLEMENTARY MATERIAL

Supplementary material is available at EJCTS online.

FUNDING

No funding has been provided.

Conflict of interest: Claudio Muneretto served as a Corcym consultant. Stefano Benussi served as an AtriCure consultant. The other authors have declared no conflict of interest.

DATA AVAILABILITY

The data underlying this article will be shared on reasonable request to the corresponding author.

Author contributions

Claudio Muneretto: Conceptualization; Resources; Supervision; Writing—review & editing. Michele D'Alonzo: Data curation; Formal analysis; Investigation; Methodology; Software; Validation; Visualization; Writing—original draft. Massimo Baudo: Formal analysis; Investigation; Methodology; Software; Validation; Visualization; Writing—review & editing. Lydia Como: Data curation; Visualization. Anna Segala: Data curation; Visualization. Francesca Zanin: Visualization; Writing—review & editing; English review. Fabrizio Rosati: Investigation; Supervision; Writing—review & editing. Stefano Benussi: Methodology; Supervision; Writing—review & editing. Lorenzo Di Bacco: Conceptualization; Formal analysis; Investigation; Supervision; Writing—original draft; Writing—review & editing

Reviewer information

European Journal of Cardio-Thoracic Surgery thanks Zinar Apaydın and the other anonymous reviewers for their contribution to the peer review process of this article.

Footnotes

REFERENCES

ABBREVIATIONS

ABBREVIATIONS
 
• AMVP

  Arrhythmic mitral valve prolapse

 
• CPB

  Cardiopulmonary bypass

 
• ECMO

  Extracorporeal membrane oxygenation

 
• GLS

  Global longitudinal strain

 
• IABP

  Intra-aortic balloon pump

 
• LAA

  Left atrial appendage

 
• LV

  Left ventricle

 
• MACEs

  Major adverse cardiac events

 
• MAD

  Mitral annular disjunction

 
• MVP

  Mitral valve prolapse

 
• MR

  Mitral regurgitation

 
• PSM

  Propensity score matching

 
• PVC

  Premature ventricular contraction

 
• SCD

  Sudden cardiac death

 
• TEE

  Transoesophageal echocardiography

 
• VA

  Ventricular arrhythmias