Outcome of mitral repair combined with Cox-maze procedure for atrial functional mitral regurgitation with heart failure with recovered ejection fraction

Abstract

Open in new tabDownload slide

OBJECTIVES

Atrial functional mitral regurgitation (AFMR) in patients with heart failure with recovered ejection fraction has received insufficient attention. This study analysed the prognosis and outcomes of mitral valve (MV) repair combined with the Cox-maze procedure.

METHODS

A prospective cohort study of patients with AFMR with left ventricular ejection fraction (LVEF) <40% was conducted. All patients received guideline-directed medical therapy. Those with recovered ejection fraction underwent MV repair combined with the Cox-maze procedure. Mortality, atrial fibrillation (AF) recurrence, mitral regurgitation (MR) and postoperative tricuspid regurgitation were assessed using the inverse probability weighting (IPW) method.

RESULTS

In total, 312 patients were enrolled in this study between 2010 and 2019, 247 of whom underwent MV repair combined with the Cox-maze procedure [full recovery (LVEF > 50%): n = 132, partial recovery (LVEF of 40–50%): n = 115]. IPW-adjusted survival of patients with LVEF ≥50% and LVEF 40–50% showed no significant difference [hazard ratio (HR): 2.18, 95% confidence interval: 0.46–10.38, P = 0.33]. However, patients with LVEF ≥50% had better IPW-adjusted long-term freedom from recurrent MR [HR: 2.44 (1.28–4.63), P = 0.0065] and AF recurrence [HR: 1.85 (1.06–3.21), P = 0.030] than those with LVEF of 40–50%.

CONCLUSIONS

MV repair combined with the Cox-maze procedure was effective and feasible in patients with severe AFMR with heart failure with recovered ejection fraction. Additionally, patients with LVEF ≥50% after guideline-directed medical therapy undergoing these combined procedures had better long-term freedom from recurrent AF and MR than those with LVEF of 40–50%.

INTRODUCTION

The occurrence of atrial functional mitral regurgitation (AFMR) in patients has not received adequate attention, despite not being a rare occurrence in clinical practice [1]. Notably, AFMR occurs in some patients with heart failure (HF) with reduced ejection fraction (HFrEF); for some, their left ventricular ejection fraction (LVEF) improves substantially after guideline-directed medical therapy (GDMT). GDMT can lead to a complete (>50%) or partial (40–50%) LVEF recovery, which is termed HF with recovered ejection fraction (HFrecEF) according to the new nomenclature [2, 3].

According to current guidelines, mitral valve (MV) repair combined with the Cox-maze procedure is indicated for patients with severe mitral regurgitation (MR) and atrial fibrillation (AF) [4, 5]. However, reports of surgical treatment outcomes for AFMR with HFrecEF are scarce [6]. In particular, the impact of MV repair combined with the Cox-maze procedure on the prognosis of these patients has not been thoroughly investigated [1]. Therefore, we aimed to assess the surgical treatment outcomes of AFMR in HFrecEF patients.

MATERIALS AND METHODS

Ethical statement

This study was approved by the ethics committee of Beijing Anzhen Hospital (No. 2022207X). Written informed consent was obtained from all the patients. The study conforms to the principles outlined in the Declaration of Helsinki. Before the analysis, all personal identifiers were replaced by codes to ensure anonymity.

Study population

We conducted a prospective observational cohort study of patients with a diagnosis of AFMR with HFrecEF who underwent MV repair combined with the Cox-maze procedure at our centre from 2010 to 2019. All patients had long-standing AF that persisted >1 year and severe MR, with no significant left ventricular remodelling (no left ventricular dilation). HFrecEF was defined as: (i) documentation of a decreased LVEF <40% at baseline, (ii) ≥10% absolute improvement in LVEF and (iii) a second measurement of LVEF ≥40%. Patients with a history of cardiac surgery and patients with MV disease secondary to rheumatic, degenerative, endocarditic, ischaemic, congenital heart or myocardial disease were excluded from the study. The medical and surgical histories of all patients were reviewed by the study coordinators to determine their eligibility. We present the following article in accordance with the STROBE statement checklist.

Surgical procedures

MV repair was performed by standard cardiopulmonary bypass. After sizing the intercommissural distance and anterior mitral leaflet area, an annuloplasty ring (Carpentier-Edwards Physio Ring II, Edwards Lifesciences, Irvine, CA, USA) was implanted with interrupted 2–0 Ethibond sutures. All patients were treated with the Cox-maze procedure, utilizing a bipolar radiofrequency clamp [7]. Tricuspid annuloplasty was subsequently performed for patients with moderate or high tricuspid regurgitation (TR) or a tricuspid annular diameter >40 mm.

Echocardiography

Standard two-dimensional and Doppler echocardiographic images were obtained by experienced cardiac sonographers in the left lateral decubitus position using a phased-array transducer in the parasternal and apical views. All echocardiography examinations were conducted in accordance with the current guidelines [8, 9]. Progression of MR and TR was defined as worsening of regurgitation by an integrative regurgitant severity grading system (jet regurgitation area, vena contracta, regurgitant fraction and effective regurgitant orifice) and graded as follows: mild = 1, moderate = 2, moderate to severe = 3, and severe = 4. Right atrial area, right ventricular area, right ventricular fractional area change (RVFAC) and tricuspid annular plane systolic excursion (TAPSE) were used to evaluate the right-sided heart structure and function. Systolic pulmonary artery pressure was measured by echocardiography using the modified Bernoulli equation on the trans-tricuspid continuous-wave Doppler signal while adding the right atrial pressure. The tricuspid valve annulus was observed in the right ventricular inflow, apical four-chamber and apical right ventricle-focused views. The tricuspid valve annulus diameter was measured in end-diastole at the level of the hinge points of the opposing tricuspid valve leaflets, as defined by the frame preceding the tricuspid valve closure.

Follow-up and outcomes

The patients were followed up at 3, 6 and 12 months, and annually thereafter. During each follow-up, the patients underwent medical history and physical evaluation, as well as 24-h Holter monitoring. All patients received postoperative antiarrhythmic drugs (AADs) and anticoagulation drugs unless contraindicated.

AADs were discontinued in patients with normal sinus rhythm 2–3 months postoperatively. Warfarin was administered for 3 months post-procedure if sinus rhythm values were stable, but was continued if AF or atrial flutter persisted. The required international normalized ratio range was between 2.0 and 3.0. A new oral anticoagulant, dabigatran, was selected for patients with recurrent AF and clotting disorders.

The primary outcome was all-cause mortality, and secondary outcomes included stroke, recurrent AF (defined as any episode of AF, atrial flutter or atrial tachycardia lasting >30 s after a 3-month blanking period), or postoperative MR with a severity > grade 3 and/or TR (regurgitation severity > grade 2).

Statistical analysis

Variables distributed abnormally were presented as medians (interquartile ranges) for continuous variables and as percentages for categorical variables. Continuous variables distributed normally were summarized as mean (standard deviation). The chi-squared test and Fisher’s test were used to test unadjusted associations between treatment variables and outcomes. T-tests and the Mann–Whitney U-test were used to assess the differences in continuous variables before adjustments. Propensity scores, which were assumed to be the probability that an individual with pretreatment characteristics X received treatment t, were acquired through the generalized boosted model. After the propensity scores were acquired, multiple treatment comparisons were performed through inverse probability of treatment weighting for causal effects. The absolute standardized mean difference was used to measure the difference between 2 univariate distributions of a single pretreatment variable. Doubly robust estimates were used if an imbalance was still present after adjusting the baseline variables, which was also defined as an augmented inverse propensity weighting (IPW). In this study, to achieve doubly robust estimations, multivariate Cox proportional hazard regression models were IPW-adjusted to determine the impact of LVEF on the outcomes. A two-sided P < 0.05 was considered statistically significant. Statistical analyses were performed using R 3.4.2 [R Development Core Team (2017)] or Stata/SE version 15 (Stata Corporation, College Station, TX, USA).

RESULTS

Baseline characteristics

From 2010 to 2019, 312 patients were diagnosed with AFMR with an LVEF <40% at the time of enrolment. Of these, 63 patients with no functional recovery of LVEF (<40%) and 2 patients whose echocardiography showed MR < grade 3 after GDMT were excluded. The 247 patients with recovered LVEF were divided into full (LVEF >50%, n = 132) and partial (LVEF of 40–50%, n = 115) recovery groups. All patients underwent scheduled primary MV repair and ablation procedures (seen in the central image). Tricuspid valve repair was performed in 58.3% of all patients. The baseline characteristics of the overall population and of the subset of the IPW-adjusted population are shown in Table 1. Although some baseline differences were noted between the 2 groups before IPW adjustment, both groups were well balanced after IPW adjustment and absolute standardized differences were usually <20%, indicating adequate matching. There were no significant differences in operative mortality (defined as death occurring within 30 days from surgery or during the same hospitalization) between the 2 groups (full recovery: 0.8%, 1/132 patients; partial recovery: 0.9%, 1/115 patients).

Long-term survival

During a mean follow-up period of 5.3 years (range, 1.4–9.2 years), 13 deaths occurred. The most frequent causes of death were heart failure (54%), bleeding (15%) and respiratory infections (15%). Five-year overall survival rates were 95% [95% confidence interval (CI): 90–97] in the total population (Fig. 1).

The overall survival of the unadjusted cohort was significantly better in patients with LVEF ≥50% [hazard ratio (HR): 4.70, 95% CI: 1.01–21.74, P = 0.029; Supplementary Material, Fig. S1A]. The multivariable-adjusted Cox model for the overall cohort showed that a longer AF duration, higher CHA2DS2-VASc score and larger effective regurgitant orifice area of MR were associated with a significantly higher risk of mortality (Fig. 2). However, there was no significant difference in IPW-adjusted survival between the 2 groups (HR: 2.18, 95% CI: 0.46–10.38, P = 0.33; Fig. 3A).

Freedom from recurrent atrial fibrillation

When examining sinus rhythm maintenance with and without AADs at follow-up, the 1- and 5-year rates of freedom from recurrent AF with and without AADs were 95% and 76% and 93% and 67%, respectively. In patients with an LVEF of 40–50%, the 1- and 5-year rates of freedom from recurrent AF with and without AADs were 90% and 68% and 90% and 62%, respectively. There were significant differences in freedom from AF recurrence on and off AADs between the 2 cohorts at follow-up (log-rank P = 0.001, seen in Supplementary Material, Fig. S2).

Similarly, in IPW-adjusted cohorts, at years 1, 2, 3, 4 and 5, the probability of freedom from AF recurrence was estimated to be 96%, 94%, 90%, 86% and 83%, respectively, in patients with an LVEF of ≥50%. AF recurrence-free survival was estimated to be 93%, 90%, 86%, 80% and 73%, respectively, at each timepoint for patients with an LVEF of 40–50%. There was a significant difference in AF recurrence-free survival between the 2 groups (HR: 1.85, 95% CI: 1.06–3.21, P = 0.030; Fig. 4).

Freedom from heart failure rehospitalization

During the follow-up period, 109 patients were readmitted for HF. The 5-year rate of freedom from HF rehospitalization was 72% (95% CI: 65–77) and 65% (95% CI: 58–72) in the total population and the IPW-adjusted cohort, respectively. The multivariable-adjusted Cox model for the overall cohort showed that longer AF duration, older age, a larger tricuspid annular diameter and a larger effective regurgitant orifice area of MR were associated with a significantly higher risk of HF rehospitalization (Fig. 5). Patients with an LVEF of <50% showed worse HF rehospitalization-free survival (HR: 1.87, 95% CI: 1.26–2.76, P = 0.025; Supplementary Material, Fig. S1B) in the unadjusted cohort. However, there was no significant difference in IPW-adjusted HF rehospitalization-free survival between the 2 groups (HR: 1.36, 95% CI: 0.91–2.03, P = 0.13; Fig. 3B).

Freedom from recurrent mitral regurgitation

Data on recurrent MR after MV repair were available for 113 patients. During follow-up, 52 of these patients developed +3 MR. The 5-year rates of freedom from 3+ MR were 87% (95% CI: 81–91) and 83% (95% CI: 78–89) in the total population and IPW-adjusted cohort, respectively. The multivariable-adjusted Cox model for the overall cohort showed that only an LVEF < 50% was associated with a significantly higher risk of recurrent +3 MR (Fig. 6). Moreover, patients with an LVEF ≥50% showed better recurrent +3 MR-free survival in the unadjusted cohort (HR: 2.32, 95% CI: 1.29–4.18, P = 0.0025; Supplementary Material, Fig. S1C) and IPW-adjusted cohorts (HR: 2.44, 95% CI: 1.28–4.63, P = 0.0065; Fig. 3C), respectively.

Based on the findings of recurrent MR, we made a three-dimensional mesh graph to demonstrate the relationship between basal LVEF, follow-up duration, and recurrent MR severity at follow-up in the entire cohort (depicted in Supplementary Material, Fig. S3). This analysis showed that as the follow-up period extended, MR severity increased and LVEF decreased.

Freedom from postoperative tricuspid regurgitation

Data on postoperative TR after MV repair were available for 189 patients. During the follow-up period, 118 of 189 patients developed 3+ TR. The 5-year rates of freedom from 3+ TR were 58% (95% CI: 51–65) and 53% (95% CI: 46–61) in the total population and the IPW-adjusted cohort, respectively. The multivariable-adjusted Cox model for the overall cohort showed that patients with an LVEF <50%, a larger right atrial area, a larger tricuspid annular diameter, and without concomitant tricuspid valvuloplasty were more likely to have +3 TR (Fig. 7).

Patients with an LVEF ≥50% showed better postoperative +3 TR-free survival in the unadjusted cohort (HR: 1.95, 95% CI: 1.34–2.85, P = 0.0004; Supplementary Material, Fig. S1D), while there was no significant difference in the IPW-adjusted cohorts (HR: 1.19, 95% CI: 0.83–1.72, P = 0.35; Fig. 3D).

Freedom from stroke

During the follow-up period, 5 embolic and 1 haemorrhagic cerebral event occurred. The 5-year rate of freedom from stroke was 97% (95% CI: 93–99) in the total population. There was no significant difference between the 2 unadjusted cohorts (HR: 2.12, 95% CI: 0.39–11.57, P = 0.39) and between the IPW-adjusted cohorts (HR: 1.59, 95% CI: 0.67–9.75, P = 0.19).

Echocardiographic follow-up

During follow-up, we found that patients with an LVEF of 40–50% showed a gradual decline in left ventricular systolic function (LVEF) and right heart function (RVFAC and TAPSE) 1 year after surgery (P < 0.01; Supplementary Material, Fig. S4B, F, and H), whereas patients with an LVEF >50% had significant reductions in these functions 5 years after surgery (P < 0.01; Supplementary Material, Fig. S4A, E, and G). However, the left ventricular end-systolic diameter significantly increased 1 year after surgery in both groups (P < 0.01; Supplementary Material, Fig. S4C and D).

DISCUSSION

The current American Heart Association/American College of Cardiology and European Society of Cardiology guidelines strongly recommend that patients with functional MR (stages C and D) and HFrEF should undergo standard GDMT for HF [10, 11]. Improvements in LVEF with GDMT can lead to a full (>50%) or partial (40–50%) recovery of LVEF in patients with HfrEF [3]. Therefore, the recognition that full recovery of LVEF in a subset of patients with functional MR and HFrEF who are treated with GDMT has led to intense interest in understanding how these patients differ from those with partial recovery of LVEF after surgical treatment.

Guidelines recommend MV surgery as a second-line treatment for severe MR due to atrial annular dilation with preserved LV systolic function (LVEF ≥50%) [12, 13]. In particular, there is a lack of clinical evaluations of surgical treatment in patients with AFMR with an LVEF <50%. To the best of our knowledge, this is the first study specifically designed to investigate AFMR with HFrecEF. By enrolling consecutive patients with HFrecEF undergoing MV repair combined with the Cox-maze procedure for AFMR, the present analysis indicated that the 5-year overall survival was 95% in the total population, and there was no significant difference in the IPW-adjusted survival between patients with an LVEF ≥50% and those with an LVEF of 40–50%. However, the patients undergoing MV repair combined with the Cox-maze procedure with an LVEF ≥50% had better long-term freedom from recurrent AF than those with an LVEF of 40–50%. Furthermore, patients with an LVEF ≥50% had better IPW-adjusted long-term freedom from recurrent MR than those with an LVEF of 40–50%. Additionally, according to echocardiographic follow-up results, while the operation was successful for patients with HFrecEF, their left heart functional reserve capacity was still greatly affected, and the risk of long-term myocardial remodelling was high. The changes in LVEF, RVFAC and TAPSE showed that in patients with an LVEF of 40–50%, the cardiac functional reserve capacity was more significantly affected. Further mechanistic research is needed to explore the underlying pathophysiology of this phenomenon.

Surgical restrictive mitral annuloplasty, which enhances leaflet coaptation by reducing annular dimensions, is the well-established gold standard approach for functional MR, based on good mid-term data from observational studies [14]. However, some studies have suggested that annulus valvuloplasty alone may not achieve satisfactory results for functional MR, particularly for MR due to left ventricular dilation [15]. Recurrence rates of up to 32.6% have been reported 12 months after the initial successful annuloplasty [16], which reflects ongoing leaflet tethering caused by continued LV remodelling, regardless of the annular size reduction [17]. However, in our report, the 5-year rate of freedom from 3+ MR was 87% in the total population after annulus valvuloplasty in patients with AFMR. Moreover, Carino et al. [18] conducted a retrospective analysis of patients with functional MR undergoing isolated ring annuloplasty and found that the cumulative incidence of recurrence of MR ≥3+ and MR ≥2+ was significantly higher in patients with MR secondary to ventricular disease than in patients with AFMR. They concluded that in patients with severe AFMR, MV repair by means of isolated ring annuloplasty might be an effective and durable treatment.

Limitations

This study had some notable limitations. First, this was a single-centre study with a small patient cohort. Second, the annuloplasty ring selected was Edwards Physio II (designed for degenerative mitral regurgitation). Moreover, an inherent limitation of the study design was the selection bias that patients with an LVEF ≥50% after GDMT seemed to have better left ventricular function. Finally, despite the application of the IPW-adjusted analysis, the interference of confounding factors still existed.

CONCLUSION

In conclusion, MV repair combined with the Cox-maze procedure was effective and feasible for patients with severe AFMR with HFrecEF. Furthermore, patients with an LVEF of ≥50% after GDMT undergoing the Cox-maze procedure and MV repair had better long-term freedom from both recurrent AF and MR than those with an LVEF of 40–50%. Larger cohorts are required to confirm these results.

SUPPLEMENTARY MATERIAL

Supplementary material is available at EJCTS online.

Funding

This work was supported by grants from the National Natural Science Foundation of China [grant number 82170311]. Professor Jiangang Wang, 1 member of this study, was awarded the funding.

Conflict of interest: none declared.

DATA AVAILABILITY

The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request.

Author contributions

Qing Ye: Conceptualization; Data curation; Investigation; Writing—original draft. Yichen Zhao: Investigation; Project administration; Writing—original draft. Chen Bai: Data curation; Formal analysis; Methodology. Kemin Liu: Data curation; Methodology. Cheng Zhao: Project administration; Validation. Yang Liu: Supervision; Validation. Yuqi Li: Formal analysis; Supervision. Jiangang Wang: Conceptualization; Supervision; Writing—review & editing.

Reviewer information

European Journal of Cardio-Thoracic Surgery thanks Tomonobu Abe, Ulrich Otto von Oppell, Gabriele Piffaretti and the other anonymous reviewer(s) for their contribution to the peer review process of this article.

REFERENCES

ABBREVIATIONS

 
• AADs

  Antiarrhythmic drugs

 
• AF

  Atrial fibrillation

 
• AFMR

  Atrial functional mitral regurgitation

 
• CI

  Confidence interval

 
• GDMT

  Guideline-directed medical therapy

 
• HFrEF

  Heart failure with reduced ejection fraction

 
• HFrecEF

  Heart failure with recovered ejection fraction

 
• HR

  Hazard ratio

 
• IPW

  Inverse probability weighting

 
• LVEF

  Left ventricular ejection fraction

 
• MV

  Mitral valve

 
• MR

  Mitral regurgitation

 
• RVFAC

  Right ventricle fractional area change

 
• TAPSE

  Tricuspid annular plane systolic excursion

 
• TR

  Tricuspid regurgitation

Author notes