Midterm results of mitral valve repair for atrial functional mitral regurgitation: a retrospective study

We retrospectively studied 40 patients who underwent MV repair for atrial functional MR between January 2011 and December 2018 at the Saitama Medical University International Medical Center (mean age, 69 ± 9 years; 68% male (n = 27)). All patients had long-standing AF that had persisted for more than 1 year, preserved LV ejection fraction (> 40%), and moderate to severe functional MR. Patients with organic valvular heart disease, including rheumatic or degenerative MV disease, a history of coronary artery disease, LV wall motion abnormality, or a history of prior cardiac surgery were excluded from the study because these conditions may confound the results. The patients’ preoperative characteristics are shown in Table 1.

The LA dimension, LV end-diastolic dimension, LV end-systolic dimension, and LV ejection fraction were measured [11]. The MR grade was evaluated using a multiparametric approach, including assessment of the Doppler-derived jet, effective regurgitant orifice area, MR volume and fraction, and pulmonary vein flow velocity pattern [12, 13]. The tricuspid regurgitation (TR) grade was also defined using a multiparametric approach [12, 13]. Pseudoprolapse of the anterior mitral leaflet (AML) was defined when there was a gap between the AML and the posterior mitral leaflet (PML) due to atriogenic PML tethering [8, 10, 14, 15]

Through median sternotomy, cardiopulmonary bypass was established using ascending aortic and bicaval cannulation. After sizing the intercommissural distance and AML area, an annuloplasty ring or band [semirigid Future annuloplasty band (n = 1), Medtronic, Minneapolis, MN; Carpentier-Edwards Physio Ring II, Edwards Lifesciences, Irvine, CA (n = 34); Cosgrove annuloplasty band (n = 3), Edwards Lifesciences; or Profile 3D (n = 2), Medtronic] was implanted with interrupted 2–0 Ethibond sutures. In the first half of this study, undersized ring annuloplasty was performed to address annular dilatation and increase the coaptation length. Once we understood the mechanisms of atrial functional MR, we performed true-sized ring annuloplasty to prevent aggravation of PML tethering. In the latter half of the present study, we used only a semi-rigid ring (Carpentier-Edwards Physio Ring II) larger than 28 mm. If AML pseudoprolapse existed, AML neochordoplasty was added. A pair of artificial neochordae with 4–0 or 5–0 polytetrafluoroethylene sutures (Gore-Tex sutures, W.L. Gore & Associates, Newark, DE) was placed at A2 from the anterolateral and posteromedial papillary muscle. Then, the length of the neochordae was adjusted after mitral ring annuloplasty until residual MR was diminished on water saline test. Concomitant tricuspid annuloplasty was performed in patients with severe TR, or mild or moderate TR with a dilated annulus (≥40 mm or > 21 mm/m² by transthoracic echocardiography) [16]. A tricuspid annuloplasty ring [MC3 ring (n = 17), Edwards Lifesciences; Physio Tricuspid annuloplasty ring (n = 10), Edwards Lifesciences; Contour 3D annuloplasty ring (n = 3), Medtronic; or a Tri-Ad Adams tricuspid annuloplasty ring (n = 1), Medtronic] was implanted with interrupted 2–0 Ethibond sutures. The indication of the maze procedure and LA plication was decided at surgeon’s discretion based on the period of atrial fibrillation and the size of LA.

Clinical follow-up examinations were completed for all patients. The mean follow-up period was 42 ± 24 months. Follow-up echocardiography was performed 1 week after surgery and every year after discharge. The mean echocardiographic follow-up period was 26 ± 24 months.

Descriptive statistics were reported as the mean ± standard deviation for continuous variables and as frequencies and percentages for categorical variables unless otherwise noted. A comparison between the groups was performed using Student’s t test or Wilcoxon-Mann-Whitney U-test for continuous variables and the chi-square or Fisher’s exact test for categorical variables. Event-free rates were estimated using Kaplan-Meier analysis. P-values < 0.05 were considered to indicate statistical significance. Statistical analyses were performed using JMP software (version 14.1.0; SAS institute, Cary, NC).

The mean LV end-systolic dimension and mean LV ejection fraction did not significantly change between the preoperative period and the follow-up period. The mean LA dimension significantly decreased from the preoperative period (60 ± 11 mm) to the follow-up period (54 ± 13 mm, p < 0.001). The mean LV end-diastolic dimension significantly decreased from the preoperative period (55 ± 7 mm) to the follow-up period (51 ± 8 mm, p < 0.001).

All patients were discharged with an MR grade no greater than mild, except one patient with moderate MR. During the follow-up period, four patients developed moderate or severe MR. Two patients with end-stage renal disease who underwent mitral ring annuloplasty alone with a 26-mm Cosgrove annuloplasty band and a 28-mm Physio Ring II developed severe MR due to partial detachment of prosthetic rings. Other patients developed moderate or severe MR due to PML tethering (Table 2). One of these patients underwent mitral ring annuloplasty with PML augmentation and basal chordae resection. Two other patients underwent mitral ring annuloplasty with AML neochordoplasty on initial surgery.

Table 3 shows the preoperative transthoracic echocardiography data of patients who developed recurrent MR (n = 5) and those without MR recurrence (n = 35). Mean LA dimensions, mean LV end-diastolic dimensions, and mean LV end-systolic dimensions were significantly larger in patients with recurrent MR than in patients without recurrent MR.

Receiver-operating characteristic curves were used to evaluate the cutoff values of LV end-diastolic dimensions as a preoperative predictor for recurrent moderate or severe MR. LV end-diastolic dimensions of 61 mm had a sensitivity of 91% and a specificity of 80% for predicting recurrent moderate or severe MR. The area under the curve for LV end-diastolic dimensions was 0.897 (Fig. 1).

Moderate or severe TR was found in 22 patients preoperatively. All patients were discharged with a TR grade no greater than mild. However, among five patients without tricuspid ring annuloplasty or sinus recovery by maze procedure, three patients developed moderate or severe recurrent TR during the follow-up period.

All patients underwent mitral ring annuloplasty. The median size of the implanted ring was 28 mm (range: 24–32 mm). As determined by intraoperative observation, mechanisms of MR included annular dilatation alone in 19 patients and annular dilatation with AML pseudoprolapse in 19 patients. Additional MV repair was performed for 22 patients, which included AML neochordoplasty in 19 patients, PML augmentation in 3, and basal chordae resection in 2. Other concomitant procedures included the maze procedure in 20 patients, tricuspid ring annuloplasty in 31, and LA plication in 9. These data are summarized in Table 4.

There were no cases of in-hospital mortality. Postoperative complication included acute renal failure requiring hemodialysis in two patients, re-exploration for bleeding in one patient, and stroke in one patient. Among the 20 patients who underwent the maze procedure, sinus rhythm was recovered in five patients (25%).

During the follow-up period, one patient experienced stroke due to infective endocarditis resulting in heart failure, seven patients required re-admission because of worsening heart failure, and six patients required permanent pacemaker implantation for sick sinus syndrome. Among five patients with recurrent MR, two patients with ring detachment required reoperation for prosthetic valve replacement 37 months and 63 months after initial operations. Seven patients died during the follow-up period because of congestive heart failure (four patients) and pneumonia (three patients) (Table 5). The rates of freedom from recurrent MR were 97.1% at 1 year, 85.4% at 3 years, and 62.3% at 5 years after surgery (Fig. 2). The rates of freedom from re-admission for heart failure were 94.9% at 1 year, 85.7% at 3 years, and 85.7% at 5 years after surgery (Fig. 3).

This study has several limitations. This was a single-center, retrospective observational study with a small number of patients. The use of multiple mitral annuloplasty rings was a major confounding variable. All control subjects underwent additional MV repair for AML pseudoprolapse or PML tethering. Thus, studies with longer follow-up periods and more statistical power are essential to validate the performance of additional MV repair for atrial functional MR. Moreover, echocardiographic studies using transesophageal echocardiography with three-dimensional morphological analysis are warranted to acquire more accurate data and to elucidate further mechanisms of recurrent MR. Finally, this study included patients with preoperative LV dilatation, which may be associated with ventriculogenic tethering. Although some argue that atriogenic and ventriculogenic should be differentiated, there is still no clear definition of atrial functional MR, and we believe that LV dilatation is one mechanism underlying the severe form of atrial functional MR. Therefore, studying the entire spectrum of atrial functional MR is crucial to elucidate the optimal surgical timing and technique.