Mitral Valve Replacement Is a Viable Alternative to Mitral Valve Repair for Ischemic Mitral Regurgitation: A Case-Matched Study

ADULT CARDIAC

Mitral Valve Replacement Is a Viable Alternative to Mitral Valve Repair for Ischemic Mitral Regurgitation: A Case-Matched Study Vincent Chan, MD, MPH, Marc Ruel, MD, MPH, and Thierry G. Mesana, MD, PhD Division of Cardiac Surgery and Department of Epidemiology and Community Medicine, University of Ottawa, Ottawa, Ontario, Canada

Background. Comparisons of mitral valve repair with mitral valve replacement for ischemic mitral regurgitation (IMR) have been limited by differences in preoperative and operative characteristics of patients undergoing these two types of surgical treatment. We performed a propensity-based, case-matched analysis to examine whether patients who undergo mitral valve repair and those who undergo mitral valve replacement for IMR have similar long-term outcomes. Methods. We compared 65 patients who underwent mitral valve replacement for IMR between 2001 and 2010 with 65 patients who underwent mitral repair during the same period on the basis of age, concomitant coronary bypass grafting, gender, left ventricular function, preoperative pulmonary hypertension, and urgency of operation. Mitral replacement involved preservation of the subvalvular apparatus. The mean study follow-up period was 2.5 ⴞ 2.1 years. Results. Two patients who underwent mitral valve repair died at 30 days postoperatively and three patients died after valve replacement. Late survival was the same

in the two groups (p ⴝ 0.4). Recurrent mitral regurgitation (MR) (grade 2ⴙ or higher) at late follow-up was observed in 15 patients (23%) after repair; however, only 1 patient (2%) had MR with a grade of more than 2ⴙ. Mitral valve repair was more commonly associated with recurrent MR (grade 2ⴙ or higher) than was mitral valve replacement (p ⴝ 0.04). Patients in both groups had similar freedom from valve-related complications and similar left ventricular function at follow-up (both p > 0.2). Conclusions. Mitral valve replacement remains a viable option for the treatment of IMR. Although mitral valve repair effectively protects against persistent or recurrent moderate-to-severe MR, mitral valve replacement provides better freedom from mild-to-moderate MR in this population, with a low incidence of valve-related complications. Notably, there was no significant difference in left ventricular function between the valve-repair and valve-replacement groups at follow-up. (Ann Thorac Surg 2011;92:1358 – 66) © 2011 by The Society of Thoracic Surgeons

I

CABG alone [5, 6], others have found that CABG with mitral valve surgery may reduce postoperative MR and cardiac-related mortality [7–9]. It is generally agreed that MR that is at least moderate to severe (grades 3⫹ or 4⫹) requires mitral valve intervention, but whether mitral valve repair or mitral valve replacement is the better type of surgical procedure remains a subject of debate [10]. In some series, mitral valve repair was found to provide better survival and to cause fewer long-term complications than mitral valve replacement. Grossi and colleagues found that at 5 years after surgery, the rate of complication-free survival of patients who underwent mitral repair for IMR was 64%, as compared with 47% for patients who underwent mitral replacement [11]. A metaanalysis of studies published between 1999 and 2009 also indicated that mitral repair conferred better short- and long-term survival than did mitral replacement [12]. The

schemic mitral regurgitation (IMR) is common and associated with poor outcomes. It is observed in up to 50% of patients presenting with acute coronary syndromes and in 19% of patients undergoing elective cardiac catheterization [1]. Approximately half of patients with IMR and recent acute coronary syndromes will develop associated heart failure [2]. Overall, patients with IMR have much worse long-term survival and functional status than do patients with coronary disease without IMR [3, 4]. The management of IMR is complex and depends in part on the severity of preoperative mitral regurgitation (MR). In patients with moderate MR (grade 2⫹), there is controversy about whether mitral valve surgery concomitant with coronary artery bypass grafting (CABG) is better than CABG alone [5– 8]. Although some advocate

Accepted for publication May 16, 2011. Presented at the Forty-seventh Annual Meeting of The Society of Thoracic Surgeons, San Diego, CA, Jan 31–Feb 2, 2011. Address correspondence to Dr Ruel, University of Ottawa Heart Institute, 3403-40 Ruskin St, Ottawa, ON, Canada, K1Y 4W7; e-mail: mruel@ ottawaheart.ca.

© 2011 by The Society of Thoracic Surgeons Published by Elsevier Inc

Dr Ruel discloses that he has financial relationships with Sanofi-Aventis, Bristol-Myers Squibb, Edwards Lifesciences, and Medtronic, Inc.

0003-4975/$36.00 doi:10.1016/j.athoracsur.2011.05.056

main benefit of mitral repair is thought to be related to the improvement in left-ventricular (LV) function that occurs with preservation of the subvalvular apparatus, which historically was often removed at the time of mitral valve replacement [13]. However, the long-term benefit of mitral repair may not apply to all patients with IMR. In several series, no short- or long-term survival benefit of mitral repair over mitral replacement was identified [14 –17]. Recurrent MR is an important problem following repair, and may be found in approximately 30% of patients at 6 months postoperatively [18]. Moreover, contemporary mitral valve replacement is done with better bioprostheses than were used in the past, as well as with preservation of the subvalvular apparatus, which may reduce the LV dysfunction that was historically associated with mitral replacement [19, 20]. The question of whether IMR should be corrected with mitral valve repair or replacement is the basis for a randomized study sponsored by the National, Heart, Lung and Blood Institute [21]. That study is currently still enrolling patients, and the clinical dilemma related to this important question is ongoing. At present, the data guiding clinicians come from published studies that are limited by differences in patients who receive valve repair versus valve replacement [15–17]. Additionally, previous attempts at matched comparisons of mitral repair and mitral replacement were significantly limited by the lack of long-term assessment of recurrent MR and, importantly, of LV function [22]. We therefore conducted a propensity-based, case-matched follow-up analysis of 130 patients who underwent either mitral repair (n ⫽ 65) or mitral replacement (n ⫽ 65) for IMR. These patients were matched on the basis of age, concomitant CABG, gender, LV function, preoperative pulmonary hypertension, and urgency of operation. We compared mitral valve repair with mitral valve replacement for the treatment of IMR with regard to survival, freedom from recurrent MR (grade 2⫹ or higher and grade 3⫹ or higher), and postoperative LV function.

Patients and Methods Ethics Approval and Funding The study was approved by the institutional review board of The University of Ottawa Heart Institute. The University of Ottawa Heart Institute Heart Valve Clinic receives unrestricted research funding from Edwards Lifesciences, Medtronic, and On-X Life Technologies.

Patient Population Between 2001 and 2010, 230 patients underwent mitral valve surgery at our institution for the correction of severe (grade 3⫹ or 4⫹), symptomatic IMR. Ischemic MR was identified by preoperative echocardiography as well as intraoperative assessment by the attending surgeon. Of these 230 patients, 69 patients underwent mitral valve replacement and 161 patients underwent mitral valve repair. Sixty-five patients who presented for surgery with

CHAN ET AL MITRAL REPAIR VERSUS REPLACEMENT FOR IMR

1359

the intent to undergo mitral valve replacement were matched on the basis of age, concomitant CABG, gender, LV function, preoperative pulmonary hypertension, and urgency of operation with 65 patients who underwent mitral valve repair. The two groups of patients constitute the basis for the present study.

Operative Characteristics All 130 patients in the study underwent mitral valve surgery via a median sternotomy, with the use of cardiopulmonary bypass (CPB), mild systemic hypothermia, and cold-blood cardioplegia. Downsizing ring annuloplasty was used for all patients who underwent mitral valve repair. Additional techniques for mitral valve repair included edge-to-edge repair in 34 patients (52%), chordal transfer in 3 patients (5%), and the use of polytetrafluroethylene chords in 2 patients (3%). Annuloplasty bands or rings used in the study included the Medtronic Futureband (Medtronic, Minneapolis, MN) in 49 patients (75%), the Duran Ancore (Medtronic) in 7 patients (10%), the Edwards Physio Ring (Baxter Healthcare Corp, Irvine, CA) in 5 patients (8%), and the Cosgrove-Edwards Annuloplasty System (Baxter) in 4 patients (6%). The mean annuloplasty ring/band size was 28.4 ⫾ 1.6 mm (median, 28 mm; range, 26 to 32 mm). Mitral valve replacement was performed with preservation of the subvalvular apparatus. The posterior mitral valve leaflet was left intact in all patients undergoing mitral replacement. In 27 of the patients undergoing mitral replacement, the anterior leaflet of the valve was partly or completely detached from the mitral annulus and divided in the middle at the 12 o’clock position, and the leftward portion of the anterior leaftlet was plicated to the anterolateral commissure with a pledgetted 4-0 polypropylene suture. The rightward portion of the anterior mitral leaflet was similarly plicated to the posteromedial commissure with a pledgetted 4-0 polypropylene suture. In the other 38 patients undergoing mitral valve replacement, the middle portion of the anterior leaflet was resected and the remaining leaflet tissue was plicated with the individual valve sutures. Prostheses implanted as mitral valve replacements included the Medtronic Hancock II, (Medtronic) in 46 patients (71%), On-X valve (On-X Life Technologies, Austin, TX) in 17 patients (26%), and Edwards Porcine (Baxter) in 2 patients (3%). The mean mitral prosthesis size was 27.9 ⫾ 2.5 mm (median, 27 mm). Concomitant CABG was performed on main coronary vessels or branches of 1.5 mm or more in diameter that displayed luminal stenosis of 70% or more on preoperative angiography.

Assessment of Preoperative and Intraoperative Mitral Regurgitation Preoperatively, all patients underwent M-mode, twodimensional, and Doppler transthoracic or transesophageal echocardiography or both for measurement of their cardiac dimensions, transvalvular pressure gradients, and severity of valvular regurgitation, as recommended by the American Society of Echocardiography (ASE) [23–25]. The mechanism of MR was classified according

ADULT CARDIAC

Ann Thorac Surg 2011;92:1358 – 66

1360

CHAN ET AL MITRAL REPAIR VERSUS REPLACEMENT FOR IMR

ADULT CARDIAC

to the system described by Carpentier [26], using both preoperative echocardiographic data and information obtained during intraoperative assessment of each patient’s mitral valve by the attending surgeon. The etiology of MR was considered to be of type I when there was at least mild annular dilation and of type IIIb when there was at least mild posterior leaflet restriction. In patients who underwent mitral valve repair, MR exceeding grade 1⫹ or a mean mitral intravalvular pressure gradient exceeding 5 mm Hg determined intraoperatively were considered indications for further mitral valve intervention. One patient had persistent MR of grade 3⫹ after initial mitral valve repair with an isolated ring annuloplasty. The valve was successfully re-repaired with the addition of edge-to-edge sutures after a return to CPB.

Follow-Up Patients were assessed annually in a dedicated valve clinic. The procedures conducted at clinic visits consisted of a history and physical examination, electrocardiogram, chest radiograph, complete blood count, serum chemistries, and determination of the international normalized ratio (INR) (when applicable). Prosthesis-related complications were recorded according to Akins and colleagues’ guidelines for reporting morbidity and mortality after cardiac valve interventions [27]. The mean postoperative follow-up period was 2.5 ⫾ 2.1 years (median, 2.1 years). Postoperative echocardiographic follow-up was also done, at 1, 3, 6, and 12 months, on all patients who had mitral valve repair. Beyond 12 months postroperatively, echocardiograms where made every 2 to 3 years or when clinically indicated. The mean echocardiographic follow-up for patients who underwent mitral valve repair was 2.1 ⫾ 2.0 years (median, 1.5 years).

Statistical Analyses Data were imported and analyzed with the Stata 10.1 statistical software program (Stata, College Station, TX). Patients who underwent mitral valve repair were compared with those who underwent mitral valve replacement with respect to demographic, echocardiographic, and operative parameters. As noted earlier, the 65 patients in the study who underwent mitral valve replacement were matched with 65 patients who underwent mitral repair on the basis of age, concomitant CABG, gender, preoperative LV function, preoperative pulmonary hypertension, and urgency of operation. These patients were matched through the calculation of a global propensity score, which was calculated on the basis of the above listed variables. Continuous data are described as mean ⫾ standard deviation. A ␹2 or Fisher’s exact test was used for categorical variables, whereas a one-way analysis of variance was used for continuous variables. A log-rank test was used to determine equality of risk factors for recurrent MR of grade 2⫹ or higher. The proportional hazards assumption was verified with Schoenfeld residuals. With 130 patients, the power to detect a hazard ratio of 1.5 with a standard deviation of 0.5 via a two-tailed test

Ann Thorac Surg 2011;92:1358 – 66

was 68%. The estimated effect size was selected on the basis of both the study data and previously published data [15–17]. Preoperative and postoperative LV function were divided into four grades based on increasing LV dysfunction: grade I, LV ejection fraction (LVEF) exceeding 50%; grade II, LVEF of 35% to 50%; grade III, LVEF of 20% to 34%; and grade IV, LVEF below 20%. Left ventricular function was compared among the groups by determining the change in LV grade following surgery. Improvement in LV function was defined as a pre- to postoperative decrease in LV grade of 1 or more. Deterioration in LV function was defined as a pre- to postoperative increase in LV grade of 1 or more. A logistic regression model was used to determine risk factors associated with a postoperative decrease in LV grade. Statistical significance was set at p ⬍ 0.05.

Results Characteristics of Patients Who Underwent Mitral Valve Repair and Those Who Underwent Mitral Valve Replacement for IMR The demographic and echocardiographic characteristics of patients who underwent mitral valve repair or replacement because of IMR are shown in Table 1. The patients’ mean age at operation was 67.7 ⫾ 9.1 years. Patients in both the mitral repair and mitral replacement groups were similar with regard to age, concomitant CABG, gender, and preoperative pulmonary hypertension. The patients in the two groups were also similar with regard to echocardiographic parameters such as preoperative left atrial (LA) size and both LV end-diastolic diameter (LVEDD) and LV end-systolic diameter (LVESD). Concomitant CABG was performed in 49 patients (75%) who underwent mitral valve repair and 56 patients (86%) who underwent mitral valve replacement. The reasons for nonperformance of concomitant CABG at the time of mitral valve surgery were percutaneous coronary intervention in 11 patients (7 who had mitral repair and 4 who had mitral replacement), previous CABG with patent bypass grafts in 8 patients (4 who had repair and 4 who had replacement), and small coronary target or distal lesions in 6 patients (5 who had repair and 1 who had replacement). Posterior mitral valve leaflet restriction (type IIIb Carpentier class of MR [26]), was the most common etiology of MR (Table 1). Among the 13 patients with type II valve prolapse who underwent mitral valve replacement, 11 had papillary muscle rupture. The remaining 2 patients with type II valve prolapse who underwent valve replacement, and 3 patients with type II valve prolapse who underwent valve repair, had elongated or torn chordae tendineae. Mitral valve repair in the latter 3 patients involved a combination of chordal transfer, the use of polytetrafluroethylene chords, and downsizing ring annuloplasty. Overall, the matching propensity scores in the mitral replacement and mitral repair groups were similar.

CHAN ET AL MITRAL REPAIR VERSUS REPLACEMENT FOR IMR

Table 1. Patient Characteristics

Variable Demographics Age (years) Female gender Concomitant CABGa 0 1 2 ⱖ3 Echocardiographic Variables Preoperative LV ejection fraction (%) Preoperative LV gradeb I II III IV Pulmonary hypertensionc Preoperative LA Size (mm) Preoperative LVEDD (mm) Preoperative LVESD (mm) Classification of MRd Type I Type II Type IIIb Operative Variables Elective surgery Thirty-day mortality Global propensity score

Mitral Valve Repair (N ⫽ 65)

Mitral Valve Replacement (N ⫽ 65)

66.9 ⫾ 8.6 20 (31%)

68.5 ⫾ 9.5 22 (34%)

0.3 0.7

16 (25%) 22 (34%) 15 (23%) 12 (18%)

10 (15%) 13 (20%) 16 (25%) 26 (40%)

0.1

37.4 ⫾ 9.3

37.0 ⫾ 12.9

0.9

pValue

1361

replacement. Causes of death included cancer (n ⫽ 5), cardiac causes (n ⫽ 5), gastrointestinal disorders (n ⫽ 3), multiorgan failure (n ⫽ 2), respiratory failure (n ⫽ 2), ruptured abdominal aortic aneurysm (n ⫽ 1), other noncardiac causes (n ⫽ 2), and unknown causes (n ⫽ 2). Long-term survival was similar in the two study groups, with a 5-year survival of 79.3 ⫾ 6.1% for patients who underwent mitral repair and of 60.6 ⫾ 8.8% for those who underwent mitral replacement, (p ⫽ 0.4) (Fig 1). Among patients with type I annular dilation or type IIIb leaflet restriction (ie, excluding patients with a type II etiology of MR), 5-year survival was also not different in the mitral repair and mitral replacement groups (79.3 ⫾ 6.1% vs 68.7 ⫾ 10.1%, respectively, p ⫽ 0.7). Five-year freedom from known cardiac-related mortality was 96.2 ⫾ 2.6% and 95.1 ⫾ 2.8% for patients who underwent mitral repair and mitral replacement, respectively (p ⫽ 0.6).

Likelihood of Recurrence of MR of Grade 2⫹ or Higher and Grade 3⫹ or Higher in Patients Undergoing Mitral Repair or Mitral Replacement

9 (14%) 20 (31%) 25 (38%) 11 (17%) 33 (51%)

18 (28%) 23 (35%) 15 (23%) 9 (14%) 36 (55%)

0.2 0.7

47.5 ⫾ 6.5

47.3 ⫾ 8.5

0.9

60.9 ⫾ 7.7

58.4 ⫾ 9.2

0.2

47.0 ⫾ 10.1

42.3 ⫾ 13.1

0.1

25 (37%) 3 (5%)e 48 (74%)

15 (23%) 13 (20%)f 48 (74%)

0.06 0.008 0.9

44 (68%) 2 (3%) 0.53 ⫾ 0.09

30 (46%) 3 (5%) 0.50 ⫾ 0.08

0.02 0.7 0.1

In follow-up, 16 patients, of whom 15 had undergone mitral valve repair and 1 had undergone mitral valve replacement, developed recurrent MR of grade 2⫹ or higher. Only 1 of the 15 patients who had undergone mitral repair had developed recurrent MR of grade 3⫹ or 4⫹ at the time of follow-up. The 1 patient who developed recurrent severe MR of grade 4⫹ after mitral replacement had endocarditis of the prosthetic valve and underwent reoperation. Of the 14 patients who developed recurrent grade 2⫹ MR after mitral valve repair, 10 had grade 4⫹ MR preoperatively and 4 others had grade 3⫹ MR preoperatively. Six-month freedom from recurrent MR of grade 2⫹ or higher was 95.0 ⫾ 3.5% after mitral valve repair and 100% following mitral valve replacement, whereas 5-year freedom from recurrent

b Test of proportions using a dichotomous variable. LV Grade I ⫽ ejection fraction ⬎50%; Grade II ⫽ ejection fraction 35–50%; Grade III ⫽ c Deejection fraction 20 –34%; Grade IV ⫽ ejection fraction ⬍20%. fined as preoperative right ventricular systolic pressure greater than 45 d Defined according to the classification of MR described by mm Hg. e Carpentier [26]. The 3 patients with type II valve prolapse who underwent mitral valve repair had elongated or torn chorf dae. Among the 13 patients with type II valve prolapse who underwent mitral valve replacement, 11 had papillary muscle rupture whereas 2 had elongated or torn chordae. a

CABG ⫽ coronary artery bypass grafting; LA ⫽ left atrium; LV ⫽ left ventricle; LVEDD ⫽ left ventricle end-diastolic diameter; LVESD ⫽ left ventricle end-systolic diameter; MR ⫽ mitral regurgitation.

Long-Term Survival of Patients With IMR Thirty-day mortality across both groups of patients in the study was 4% (5 of 130 patients), and was not significantly different in the two groups (Table 1). Twenty-two late deaths were observed during follow-up, of which 10 occurred after mitral valve repair and 12 after mitral valve

Fig 1. Long-term survival after mitral repair or replacement. This figure represents all-cause mortality for mitral valve surgery for ischemic mitral regurgitation (I MR). In total, 22 deaths were observed in this cohort: 10 after mitral valve repair and 12 after mitral valve replacement. Causes of death included cancer (n ⫽ 5), cardiac causes (n ⫽ 5), gastrointestinal disorders (n ⫽ 3), multiorgan failure (n ⫽ 2), respiratory failure (n ⫽ 2), ruptured abdominal aortic aneurysm (n ⫽ 1), other noncardiac causes (n ⫽ 2), and unknown causes (n ⫽ 2). Concomitant coronary artery bypass grafting (CABG) was performed in 46 patients who underwent mitral valve repair and 55 patients who underwent mitral valve replacement.

ADULT CARDIAC

Ann Thorac Surg 2011;92:1358 – 66

1362

CHAN ET AL MITRAL REPAIR VERSUS REPLACEMENT FOR IMR

Ann Thorac Surg 2011;92:1358 – 66

ADULT CARDIAC

Fig 2. Freedom from recurrent mitral regurgitation (MR) after mitral repair or replacement. Recurrent MR was determined through postoperative echocardiographic assessments. All patients underwent transthoracic echocardiography following repair according to clinic guidelines. In patients with multiple echocardiographic assessments, the earliest date in which recurrent MR developed was used for analysis. However, if recurrent MR was not observed, the most recent date of echocardiography was utilized. Figure 2A describes recurrent MR (ⱖ2⫹) whereas Figure 2B describes recurrent MR 3⫹ or 4⫹.

MR of grade 2⫹ or higher was 41.4 ⫾ 14.8% following mitral valve repair and 85.7 ⫾ 13.2% after mitral valve replacement (p ⫽ 0.04) (Fig 2). Five-year freedom from recurrent MR of grade 3⫹ or higher was 90.9 ⫾ 8.7% after mitral repair and 85.7 ⫾ 13.2% after mitral replacement (p ⫽ 0.9). The number of grafts performed at mitral valve surgery was inversely associated with the recurrence of MR of grade 2⫹ or higher (hazard ratio [HR] 0.5 per additional graft, 95% confidence interval [CI] 0.3 to 0.9, p ⫽ 0.02) (Table 2).

Valve-Related Complications and Late LV function With Mitral Valve Repair and Mitral Valve Replacement Two mitral valve-related complications occurred in patients in the study who underwent valve replacement. One was a

Table 2. Risk Factors Associated With Recurrent MR (ⱖ2⫹) Risk Factor

Lack of Association of Etiology of MR and of Size of Annuloplasty Band or Ring With Development of Recurrent MR of Grade 2⫹ or Higher in Patients Undergoing Mitral Valve Repair for IMR Patients who developed recurrent MR of grade 2⫹ or higher after mitral valve repair were similar in terms of age, gender, and preoperative LV function to those who underwent repair without developing MR (Table 3). The etiology of IMR, whether type I or type IIIb according to the Carpentier classification, was not associated with recurrent MR of grade 2⫹ or higher (p ⬎ 0.6 for both types of etiology). Nor was the ring or band size used in annuloplasty associated with recurrent MR of grade 2⫹ or higher (p ⫽ 0.9).

Hazard Ratio 95% CI p-Value

Significant Covariates Number of coronary grafts (per additional graft) Mitral valve repair Non-Significant Covariates Age (per increasing year) Female gender Preoperative LV function (per increasing grade)a

0.5

0.3–0.9

0.02

8.0

1.1–62.4

0.04

1.0 0.6 1.2

0.9–1.0 0.1–2.5 0.6–2.4

0.5 0.4 0.6

a LV Grade I ⫽ ejection fraction ⬎50%; Grade II ⫽ ejection fraction 35–50%; Grade III ⫽ ejection fraction 20 –34%; Grade IV ⫽ ejection fraction ⬍20%.

CI ⫽ confidence interval; regurgitation.

LV ⫽ left ventricle;

MR ⫽ mitral

Ann Thorac Surg 2011;92:1358 – 66

Comment In this study we compared 65 patients who underwent mitral valve repair with 65 patients who underwent mitral valve replacement for the treatment of IMR. These patients were matched according to age, concomitant CABG, gender, preoperative LV function, preoperative pulmonary hypertension, and urgency of operation. The study was done to compare mitral valve repair with mitral valve replacement in terms of long-term postoperative survival, freedom from recurrent MR, and postoperative LV function.

Table 3. Risk Factors for the Development of Recurrent MR (ⱖ2⫹) in Mitral Valve Repair Patients

Age (per increasing years) Concomitant CABGa Female gender Preoperative LV gradeb Classification of MRc Type I Type IIIb Annuloplasty size (per increasing mm)

Hazard Ratio

95% CI

p-Value

0.99 0.3 0.5 1.1

0.92–1.06 0.1–1.1 0.1–2.4 0.5–2.3

0.7 0.07 0.4 0.7

1.0 1.4 1.0

0.3–3.3 0.4–5.4 0.7–1.5

0.9 0.6 0.9

b Entered into model as an ordinal variable. LV Grade I ⫽ ejection fraction ⬎50%; Grade II ⫽ ejection fraction 35–50%; Grade III ⫽ ejection c Defined acfraction 20 –34%; Grade IV ⫽ ejection fraction ⬍20%. cording to the classification of MR described by Carpentier [26].

a

CABG ⫽ coronary artery bypass grafting; CI ⫽ confidence interval; LV ⫽ left ventricle; MR ⫽ mitral regurgitation.

Table 4. Change in LV Function AfterMitral Valve Repair or Replacementa

Type of Surgery Mitral valve repair Mitral valve replacement p Value

LV Improvement (N ⫽ 37)

No Change in LV LV Function Deterioration (N ⫽ 46) (N ⫽ 47)

16 21

23 23

0.5

26 21

0.9

0.6

Improvement in LV function was defined as an increase of ⱖ1 grade postoperatively compared to preoperatively. Similarly, deterioration of LV function was defined as a decrease of ⱖ1 grade postoperatively compared to preoperatively.

a

LV Grade I ⫽ ejection fraction ⬎50%; Grade II ⫽ ejection fraction 35–50%; Grade III ⫽ ejection fraction 20 –34%; Grade IV ⫽ ejection fraction ⬍20%. LV ⫽ left ventricle.

The short-term and long-term mortalities of patients in the study compare with those in previously published series. The published operative mortality of patients undergoing CABG with concomitant mitral valve repair or replacement for IMR ranges from 4% to 14% [7, 14 –16, 28], and their reported unadjusted 5-year survival ranges from 44% to 64% [29 –30]. Although some investigators have found that mitral valve repair is associated with better survival than is mitral valve replacement [31–32], others have found no difference [14 –16]. Gillinov and associates surmised that in patients with very severe LV dysfunction, the surgical approach to the mitral valve is less important given the poor overall survival in this population [22]. In the present study we observed no difference in overall survival in the mitral valve repair and mitral valve replacement groups. We also found that the freedom from cardiac-related death after either valve repair or replacement was greater than 90%. Lower values for Table 5. Risk Factors Associated With a Decrease in Postoperative LV Function After Mitral Valve Repair or Replacement Risk Factor Significant covariates Concomitant CABGa Non-Significant covariates Age (per increasing year) Female gender Preoperative LV function (per increasing grade)b Mitral valve repair

Odds Ratio

95% CI

p-value

0.2

0.04–0.8

0.02

1.02 0.5 1.4

0.95–1.10 0.1–1.9 0.7–2.9

0.5 0.3 0.3

1.0

0.2–3.8

0.9

a b Entered into logistic regression model as a dichotomous variable. Improvement in LV function was defined as a decrease of ⱖ1 grade postoperatively compared to preoperatively. Similarly, worsening of LV function was defined as an increase of ⱖ1 grade postoperatively compared to preoperatively.

LV Grade I ⫽ ejection fraction ⬎50%; Grade II ⫽ ejection fraction 35–50%; Grade III ⫽ ejection fraction 20 –34%; Grade IV ⫽ ejection fraction ⬍20%. CABG ⫽ coronary artery bypass grafting; val; LV ⫽ left ventricle.

CI ⫽ condidence inter-

ADULT CARDIAC

case of endocarditis (mentioned above) in a prosthetic mechanical mitral valve, requiring reoperation at 24 months after initial mitral valve replacement surgery. The other was a case of structural deterioration of a porcine valve, which presented as severe mitral stenosis that required reoperation with the insertion of a mechanical prosthesis at 47.5 months after initial valve replacement. No hemorrhage or thromboembolism was observed during the study period in patients who underwent valve replacement. One case of peripheral embolism occurred at 73.2 months after mitral valve repair in a patient who was being treated with warfarin for chronic atrial fibrillation, and another patient in the mitral repair group developed MR of grade 4⫹ at 5.5 months after valve surgery, as a result of bacterial endocarditis. This patient subsequently underwent successful mitral valve replacement. Seven-year freedom from valve-related complications was 80.0 ⫾ 17.9% after mitral valve repair and 84.3 ⫾ 9.2% after mitral valve replacement (p ⫽ 0.2). The proportion of patients whose LV function improved or deteriorated postoperatively as compared with its preoperative level did not differ with either mitral valve repair or replacement (both p ⬎ 0.5) (Table 4). Nor was mitral valve repair associated with a decrease in postoperative LV grade (odds ratio [OR] 1.0, 95% CI 0.2 to 3.8, p ⫽ 0.9), although concomitant CABG was associated with such a decrease (OR 0.2, 95% CI 0.04 to 0.8, p ⫽ 0.02) (Table 5).

Risk Factor

1363

CHAN ET AL MITRAL REPAIR VERSUS REPLACEMENT FOR IMR

1364

CHAN ET AL MITRAL REPAIR VERSUS REPLACEMENT FOR IMR

ADULT CARDIAC

freedom from cardiac-related death have been reported in other series [7–9]. This difference may reflect the relatively high proportion of noncardiac causes of death in our study. Four percent of our patients died of cancer and 6% died of pulmonary, gastrointestinal, or vascular causes. In this study, mitral valve repair was associated with higher rates of recurrent MR of grade 2⫹ or more than was mitral valve replacement. Six-month freedom from recurrent MR of grade 2⫹ or higher was 95.0 ⫾ 3.5% after repair, which compares favorably with that in other series [18]. Although only one patient in our study developed MR exceeding grade 2⫹ after mitral valve repair, the 5-year freedom from recurrent MR of grade 2⫹ or higher was 41%. The recurrence of MR may be related to incomplete LV reverse remodeling [31]. Preoperative LVEDD enlargement of more than 65 to 70 mm is associated with poor LV remodeling and recurrent MR [32, 33]. In patients who developed recurrent MR of grade 2⫹ or more after repair, we did not find that preoperative LV size was a significant risk factor. However, the number of coronary grafts done in our study was an important determinant of recurrent MR of grade 2⫹ or higher. These observations suggest that the degree of potential LV recovery as a result of revascularization is a more important predictor of recurrent MR of grade 2⫹ or higher than is absolutely measured preoperative LV size. Not surprisingly, a small annuloplasty band or ring size alone failed to protect against recurrent MR, which is in keeping with observations made by others [18]. Another important observation in our study was that mitral valve replacement was associated with few valverelated complications. Despite the single case of prosthetic-valve endocarditis and single case of structural deterioration of a porcine bioprosthesis, the overall rate of occurrence of endocarditis and structural valve deterioration in the study was lower than that in other published series [34, 35]. Only 26% of the patients in the study received a mechanical prosthesis, which may partly explain the lack of hemorrhagic or thromboembolic events. Low rates of thromboembolism have been reported with the On-X valve, which was the only mechanical valve used in the study [36]. The case of peripheral embolism in the mitral valve repair group occurred in a patient with chronic atrial fibrillation, which is a well-established risk factor for thromboembolism [37].

Study Limitations Patients who underwent mitral valve repair were matched with those who underwent mitral valve replacement on the basis of a calculated propensity score. The two groups may have differed with regard to important risk factors not included in the propensity score. Matching also fails to completely eliminate important but statistically nonsignificant differences between groups. Although there was no difference between the two groups in the study with regard to concomitant CABG, there was a trend toward more coronary grafts in patients who underwent mitral valve replacement than in those who had mitral valve repair. Because the study compared mitral valve repair with mitral valve replacement in patients with severe preoperative MR,

Ann Thorac Surg 2011;92:1358 – 66

no comparisons were made involving patients with moderate MR who underwent CABG without mitral valve surgery. Also, because noninvasive measurements of myocardial viability were not available preoperatively, the benefit of CABG in patients with LV dysfunction remains undetermined. This last point is not only relevant for postoperative LV function, but also for survival and the development of recurrent MR. Few valve-related complications were observed in the study, but more complications would have been observed with longer follow-up, larger numbers of patients, or both. The power to detect a difference in overall survival or cardiac-related survival was also limited and can similarly be increased with longer follow-up.

Conclusions Mitral valve replacement remains a viable option for treating IMR. Although mitral valve repair effectively protects against persistent or recurrent moderate to severe MR, mitral replacement provides better freedom from mild to moderate MR in the population with IMR, with a low incidence of valve-related complications. Notably, the mitral replacement and mitral repair groups showed no significant difference in LV function at follow-up.

References 1. Hickey MS, Smith LR, Muhlbaier LH, et al. Current prognosis of ischemic mitral regurgitation: implications for future management. Circulation 1988;78:151–9. 2. Trichon BH, Felker GM, Shaw LK, et al. Relation of frequency and severity of mitral regurgitation to survival among patients with left ventricular systolic dysfunction and heart failure. Am J Cardiol 2003;91:538 – 43. 3. Bursi F, Enriquez-Sarano M, Nkomo VT, et al. Heart failure and death after myocardial infarction in the community: the emerging role of mitral regurgitation. Circulation 2005;111: 295–301. 4. Lamas GA, Mitchell GF, Flaker GC, et al. Clinical significance of mitral regurgitation after acute myocardial infarction. Circulation 1997;96:827–33. 5. Ryden T, Bech-Hanssen O, Brandrup-Wognsen G, Nilsson F, Svensson S, Jeppsson A. The importance of grade 2 ischemic mitral regurgitation in coronary artery bypass grafting. Eur J Cardiothorac Surg 2001;20:276 – 81. 6. Duarte IG, Shen Y, MacDonald MJ, Jones EL, Craver JM, Guyton RA. Treatment of moderate mitral regurgitation and coronary disease by coronary bypass alone: late results. Ann Thorac Surg 1999;68:426 –30. 7. Aklog L, Filsoufi F, Flores KQ, et al. Does coronary artery bypass grafting alone correct moderate ischemic mitral regurgitation? Circulation 2001;104:168 –75. 8. Lam BK, Gillinov AM, Blackstone EH, et al. Importance of moderate ischemic mitral regurgitation. Ann Thorac Surg 2005;79:462–70. 9. Fattouch K, Sampognaro R, Speziale G, et al. Impact of moderate ischemic mitral regurgitation after isolated coronary artery bypass grafting. Ann Thorac Surg 2010;90:1187–94. 10. Bonow RO, Carabello BA, Chatterjee K, et al. 2008 focused update incorporated into the ACC/AHA 2006 guidelines for the management of patients with valvular heart disease: a report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines (Writing Committee to revise the 1998 guidelines for the management of patients with valvular heart disease). J Am Coll Cardiol 2008;52:676 – 85. 11. Grossi EA, Goldberg JD, LaPietra A, et al. Ischemic mitral valve reconstruction and replacement: comparison of longterm survival and complications. J Thorac Cardiovasc Surg 2001;122:1107–24.

12. Vassileva CM, Boley T, Markwell S, Hazelrigg S. Metaanalysis of short-term and long-term survival following repair versus replacement for ischemic mitral regurgitation. Eur J Cardiothorac Surg 2011;39:295–303. 13. David TE, Burns RJ, Bacchus CM, Druck MN. Mitral valve replacement for mitral regurgitation with and without preservation of chordae tendineae. J Thorac Cardiovasc Surg 1984;88:718 –25. 14. Calafiore A, Di Mauro M, Gallina S, et al. Mitral valve surgery for chronic ischemic mitral regurgitation. Ann Thorac Surg 2004;77:1989 –97. 15. Cohn LH, Rizzo RJ, Adams DH, et al. The effect of pathophysiology on the surgical treatment of ischemic mitral regurgitation: operative and late risks of repair versus replacement. Eur J Cardiothorac Surg 1995;9:568 –74. 16. Magne J, Girerd N, Sénéchal M, et al. Mitral repair versus replacement for ischemic mitral regurgitation: comparison of short-term and long-term survival. Circulation 2009; 120(11 Suppl):S104 –11. 17. Al-Radi OO, Austin PC, Tu JV, David TE, Yau TM. Mitral repair versus replacement for ischemic mitral regurgitation. Ann Thorac Surg 2005;79:1260 –7. 18. McGee EC, Gillinov AM, Blackstone EH, et al. Recurrent mitral regurgitation after annuloplasty for functional ischemic mitral regurgitation. J Thorac Cardiovasc Surg 2004;128:916 –24. 19. DeAnda AJ, Komeda M, Nikolic SD, Daughters GT, Ingels NB, Miller DC. Left ventricular function, twist, and recoil after mitral valve replacement. Circulation 1995;92:II458 – 66. 20. Ruel M, Kulik A, Rubens FD, et al. Late incidence and determinants of reoperation in patients with prosthetic heart valves. Eur J Cardiothorac Surg 2004;25:364 –70. 21. Comparing the Effectiveness of Repairing Versus Replacing the Heart’s Mitral Valve in People With Severe Chronic Ischemic Mitral Regurgitation Clinical Trials Registration. Available at http://clinicaltrials.gov/ct2/show/NCT00807040. Accessed January 20, 2011. 22. Gillinov AM, Wieryp PN, Blackstone EH, et al. Is repair preferable to replacement for ischemic mitral regurgitation? J Thorac Cardiovasc Surg 2001;122:1125– 41. 23. Zoghbi WA, Enriquez-Sarano M, Foster E, et al. Recommendations for evaluation of the severity of native valvular regurgitation with two-dimensional and Doppler echocardiography. J Am Soc Echocardiogr 2003;16:777– 802. 24. Lang RM, Bierig M, Devereux RB, et al. Recommendations for chamber quantification: a report from the American Society of Echocardiography’s Guidelines and Standards Committee and the Chamber Quantification Writing Group, developed in conjunction with the European Association of Echocardiography, a branch of the European Society of Cardiology. J Am Soc Echocardiogr 2005;18:1440 – 63. 25. Quinones MA, Otto CM, Stoddard M, Waggoner A, Zoghbi WA. Recommendations for quantification of Doppler echo-

CHAN ET AL MITRAL REPAIR VERSUS REPLACEMENT FOR IMR

26. 27. 28.

29.

30. 31. 32. 33.

34. 35. 36.

37.

1365

cardiography: a report from the Doppler Quantification Task Force of the Nomenclature and Standards Committee of the American Society of Echocardiography. J Am Soc Echocardiogr 2002;15:167– 84. Carpentier A. Cardiac valve surgery—the “French correction.” J Thorac Cardiovasc Surg 1983;86:323–37. Akins CW, Miller DC, Turina MI, et al. Guidelines for reporting mortality and morbidity after cardiac valve interventions. Ann Thorac Surg 2008;85:1490 –5. Diodato MD, Moon MR, Pasque MK, et al. Repair of ischemic mitral regurgitation does not increase mortality or improve long-term survival in patients undergoing coronary artery revascularization: A propensity analysis. Ann Thorac Surg 2004;78:794 –9. Glower DD, Tuttle RH, Shaw LK, Orozco RE, Rankin JS. Patient survival characteristics after routine mitral valve repair in ischemic mitral regurgitation. J Thorac Cardiovasc Surg 2005;129:860 – 8. Gillinov AM, Blackstone EH, Rajeswaran J, et al. Ischemic versus degenerative mitral regurgitation: does etiology affect survival? Ann Thorac Surg 2005;80:811–9. Bishay ES, McCarthy PM, Cosgrove DM, et al. Mitral valve surgery in patients with severe left ventricular dysfunction. Eur J Cardiothorac Surg 2000;17:213–21. Dahlberg PS, Orszulak TA, Mullany CJ, et al. Late outcome of mitral valve surgery for patients with coronary artery disease. Ann Thorac Surg 2003;76:1539 – 48. Grossi EA, Woo YJ, Patel N, et al. Outcomes of coronary artery bypass grafting and reduction annuloplasty for functional ischemic mitral regurgitation: a prospective multicenter study (Randomized Evaluation of a Surgical Treatment for Off-Pump Repair of the Mitral Valve). J Thorac Cardiovasc Surg 2011;141:91–7. Ikonomidis JS, Kratz JM, Crumbley AJ 3d, et al. Twenty-year experience with the St Jude Medical mechanical valve prosthesis. J Thorac Cardiovasc Surg 2003;126:2022–31. Chan V, Kulik A, Tran A, et al. Long-term clinical and hemodynamic performance of the Hancock II versus the Perimount aortic bioprostheses. Circulation 2010;122(11 Suppl):S10 – 6. Chan V, Jamieson WR, Lam BK, et al. Influence of the On-X mechanical prosthesis on intermediate-term major thromboembolism and hemorrhage: a prospective multicenter study. J Thorac Cardiovasc Surg 2010;140:1053– 8. Fuster V, Ryden LE, Cannom DS, et al. ACC/AHA/ESC 2006 Guidelines for the Management of Patients With Atrial Fibrillation A Report of the American College of Cardiology/ American Heart Association Task Force on Practice Guidelines and the European Society of Cardiology Committee for Practice Guidelines (Writing Committee to Revise the 2001 Guidelines for the Management of Patients With Atrial Fibrillation). J Am Coll Cardiol 2006;48:854 –906.

DISCUSSION DR ROBERT DION (Genk, Belgium): A very nice presentation. Of course, nobody can ever prove that replacement is inferior to repair without a prospective randomized study. You present 25% recurrence of at least moderate MR at 2 years after valve repair but, you have used, as the Cleveland Clinic in its original paper by McGee,a variety of repair techniques and devices: a band, an incomplete band, a complete ring, . . . You do not precise either how restrictive the mitral valve annuloplasty was. And my problem with that type of study with inhomogeneous techniques and devices resulting in a 25% recurrence of moderate MR at 2 years, is that you certainly do not favor the restrictive mitral annuloplasty! In Leiden we have published 13% of recurrence after 4 years because we use, all of us, the

same strategy. Therefore, my first question: how do you achieve the restriction of the mitral annulus, what is your endpoint, do you measure the leaflet coaptation, and when do you decide to use an incomplete band versus a rigid complete ring? And my second question is about the outcome of the patient. In your series, the mean left ventricular end-diastolic diameter was 60 mm. In our experience, these LV undergo a reverse remodeling after a successful and durable restrictive annuloplasty. The same should happen after a mitral valve replacement, supposedly by using a bioprosthesis. What happens then after 5 years in a patient of, let us say, 50 years who is optimally revascularized and who has ‘reversely remodeled’: a reoperation for a degenerated bioprothesis in mitral position?

ADULT CARDIAC

Ann Thorac Surg 2011;92:1358 – 66

1366

CHAN ET AL MITRAL REPAIR VERSUS REPLACEMENT FOR IMR

ADULT CARDIAC

DR CHAN: Thank you for your questions, Dr. Dion. Your first question relates to the nature of our mitral valve repair techniques, and also the impact of recurrent 2-plus MR on outcomes. Seventy-five percent of the patients in this study received the Medtronic Futureband, which is a rigid band. The paper that you quoted from the Cleveland Clinic reported a 28% rate of recurrent MR at 6 months. The focus of that paper was to compare different annuloplasty bands, some of which were softer that the rigid Medtronic Futureband. In this study, few patients received a complete O-ring like the Physio-ring or a soft Duran band. At our institution, if excessive tethering of the posterior leaflet was observed intra-operatively, the Alfieri technique was utilized. Thirty-four patients, or over half of our repair patients, received an edge-to-edge repair. The rate of recurrent MR greater than or equal to 2-plus is not well-defined in the literature. The paper that you quoted from the Cleveland Clinic addressed recurrent 3- and 4-plus MR. Our experience is that recurrent moderate MR may not impact clinical endpoints such as New York Heart Association functional class. This data was not included in today‘s presentation. Your other question related to mitral valve replacement. Prior to surgery, the operating surgeon would discuss the advantages and disadvantages of different mitral valve prostheses as part of the preoperative consent process. In younger patients, a mechanical valve may be preferable. In our series, 17 patients received the On-X valve, whereas the majority received a tissue valve. This is a reflection of the patient demographics of this study. One patient developed premature structural deterioration of a porcine mitral valve approximately 4 years following initial mitral surgery. That patient required reoperation for mitral stenosis rather than insufficiency. These data highlight the tradeoff between freedom from valve-related complications versus the rate of recurrent mitral insufficiency, when considering mitral valve repair versus replacement. DR SCOTT MCCLURE (London, Ontario, Canada): Vince, good job as per usual. Just a quick question. To buy into these results, obviously, as you know, with a study such as this, we have to believe in the propensity model. And my question is more to do with your model. If I understood your study, you were able to match 65 of 69 patients to a pool of only 131. Is that correct? DR CHAN: Between 2001 and 2010, 230 patients underwent mitral valve surgery for ischemic mitral regurgitation at our institution. Of these, 161 underwent mitral valve repair, and 69 patients underwent replacement. The 65 replacement patients included in this study represent individuals that presented to the OR with the intention of receiving a prosthetic mitral valve. This is different from individuals for which mitral valve repair was attempted, prolonged cardiopulmonary bypass was required, and eventually, received mitral valve replacement. Several aforementioned preoperative variables were used to calculate a propensity score in order to facilitate matching with 65 mitral valve repair patients. Another way to analyze these data would have been to use a regression model. We did repeat the analysis with a proportional hazards model, and similar results were achieved. We ultimately decided to present the data using a propensity score in order to simplify our results. DR MCCLURE: And with regards to the preoperative variables you used, like ejection fraction, did you simply dichotomize the variable above and below a particular threshold? How liberal, or put

Ann Thorac Surg 2011;92:1358 – 66

another way, how strict were you? Because when you look at the baseline demographics, you “means” were similar for left ventricular diameter as an average across the whole group; but for the individual matching, how strict were you? And with regards to the grafts, did you simply dichotomize the variable: Yes, you had a graft, or no, you didn’t? Or did you actually model them so you matched individuals patients based on grafts to areas with an ischemic burden? Because if we don’t have confidence in the model, then I just don’t know that you can be as confident in the results. DR CHAN: I appreciate your comments. Ejection fraction was categorized into four strata of increasing left ventricle dysfunction. This is consistent with the outcomes papers published from our institution. One benefit of this categorization is that it allows minor measurement and recording errors within each strata. Concomitant coronary artery bypass grafting was used as a dichotomous variable for all regression models. DR ZIVOJIN JONJEV (Sremska Kamenica, Serbia): Repair or replacement is a very simple question, but a very complex problem. At the present time we have a lot of information about Ischemic Mitral Regurgitation. One of the most important things is the extent of remodeling of the left ventricle and it’s ejection fraction. You mentioned that you have a significant number of patients with ejection fraction lower than 35%. In most of the recently published literature the cutoff point in Ischemic Mitral Regurgitation is ejection fraction lower than 30%, and those patients have completely different treatment, prognosis, and survival curves. So those patients should be observed and presented separately. The second issue is decision-making process in your team. How do you make patient selection for repair or replacement? Do you respect sphericity index, mitral valve tenting, and ejection fraction together, or do you have some other criteria? And finally, your survival and follow up curves. Significant difference between repair and replacement is detectable 5 or 6 years after surgery. Your follow up lasts just 5 or 6 years. So maybe you should extend your research in the future for a longer period of time. DR CHAN: Thank you kindly for your comments. Preoperative ejection fraction is an important predictor of long-term outcomes. The intention of this study was to investigate repair versus replacement in a “real-world” population of ischemic MR. By matching patients, we had a similar proportion of LV dysfunction between groups. In this study, no assessments of preoperative myocardial viability were made. The extent of myocardial viability would determine the degree of LV recovery following coronary artery bypass grafting. As such, preoperative LV size alone may not completely predict remodeling following surgery. The decision as to whether or not a valve was repairable was ultimately made by the operating surgeon. Intra-operative echocardiographic measurements such as posterior angle, posterior leaflet length, coaptation distance were not presented in this data today. I would argue that these echocardiographic parameters are important for predicting recurrent MR in the early postoperative period. In this series, the 6-month freedom from recurrent MR more than 2 plus was 95%. In my opinion, variables that portend poor durability of mitral valve repair in the long term are not well described. And I absolutely agree that these patients should be followed further into the future. Although we report no difference in survival between groups, a difference may be observed with longer follow-up.