- Email: [email protected]
Mitral valve repair for ischemic mitral insufficiency
Mitral Valve Repair for Ischemic Mitral Insufficiency William G. Hendren, MD, James J. Nemec, MD, Bruce W. Lytle, MD, Floyd D. Loop, MD, Paul C. Taylor, MD, Robert W. Stewart, MD, and Delos M. Cosgrove 111, MD Department of Thoracic and Cardiovascular Surgery and Department of Cardiology, The Cleveland Clinic Foundation, Cleveland, Ohio
(lo),
Over a 5-year period, 1,292 patients had operation on their native mitral valves. Ischemia was the cause of mitral insufficiency in 84 patients (6.5%). Sixty-five patients (77.4%)had mitral valve repair. Mean age was 66 & 10 years; 35 patients (53.8%)were women. Mean degree of preoperative insufficiency was 3.2 & 0.7; mean preoperative New York Heart Association functional class was 3.3 & 0.7. Eleven patients (16.9%)had acute and 54 (83.1%)had chronic mitral insufficiency. Valve prolapse was present in 26 patients (40%). Restrictive leaflet motion secondary to regional or global left ventricular dilatation occurred in 39 patients (60%).All patients had associated myocardial revascularization followed by transatrial valvuloplasty. Multiple techniques were employed to achieve valve competence: leaflet resection (31, chordal shortening (15), papillary muscle reimplantation
papillary muscle shortening (3), and annuloplasty (63). There were six (9.2%)hospital deaths (acute, 9.1%; chronic, 9.3% [not significant]; prolapse, 11.5%; restrictive, 7.7% [not significant]). The mean degree of postoperative mitral insufficiency was 0.6 2 0.8 in 51 patients. At a mean follow-up of 3.1 f 1.6 years, patient survival was 96% for patients with valve prolapse and 48% for those with restrictive leaflet motion ( p = 0.02). New York Heart Association functional class was improved in all groups. Ischemic mitral insufficiency is an uncommon cause of mitral valve disease that is amenable to repair in the majority of cases of both acute and chronic onset. The operative mortality is low, and operation is associated with superior survival in patients with valve prolapse.
I
Patients and Methods
schemic mitral regurgitation remains one of the most challenging management problems in cardiac surgery and is associated with suboptimal results. The majority of patients with ischemic regurgitation are treated medically, but patients with severe regurgitation were found to benefit from surgical treatment [ 1-31. However, the operative morbidity and mortality associated with operation for ischemic mitral regurgitation are higher than those for other forms of mitral valve insufficiency. In the last decade, the operative approach to moderate or severe ischemic regurgitation has evolved from myocardial revascularization to surgical revascularization combined with mitral valvuloplasty. Recognized advantages of mitral valve repair compared with prosthetic valve replacement are lower operative mortality, better preservation of ventricular function, reduced valverelated complications, decreased requirement for anticoagulation therapy, improved long-term survival, and decreased cost [4]. To evaluate the efficacy of mitral valve repair for ischemic regurgitation, our experience over a 4X-year period was reviewed. Presented at the Twenty-seventh Annual Meeting of The Society of Thoracic Surgeons, San Francisco, CA, Feb 1E-20, 1991. Address reprint requests to Dr Cosgrove, Department of Thoracic and Cardiovascular Surgery, The Cleveland Clinic Foundation, 9500 Euclid Ave, Cleveland, OH 44195-5066.
0 1991 by The Society of Thoracic Surgeons
(Ann Thoruc Surg 1991;52:1246-52)
All patients undergoing operation for moderate to severe ischemic mitral insufficiency were included. Ischemic mitral regurgitation was defined as valve insufficiency secondary to myocardial infarction with angiographic or echocardiographic documentation of left ventricular wall motion abnormality in a region perfused by a critically obstructed coronary artery. Patients were excluded if there was any macroscopic or histologic evidence of primary mitral valve pathology. Ischemic regurgitation was considered acute if the interval between myocardial infarction and operation was less than 4 weeks and chronic if the interval was greater. Between January 1985 and June 1989, 1,292 patients underwent valve procedures on their native mitral valves at The Cleveland Clinic Foundation. Eighty-four patients (6.5%)had ischemic mitral insufficiency (Fig 1).Nineteen patients (22.6%) underwent revascularization and valve replacement; 65 (77.4%) underwent revascularization combined with mitral valve repair and form the basis of this report. Eleven patients (16.9%) had acute ischemic insufficiency and 54 (83.1%) had chronic. There were 30 men and 35 women. Patients ranged in age from 47 to 78 years (mean age, 66.0 2 8.1 years). The mean interval from myocardial infarction to operation was 77.1 2 134.6 weeks for the entire group and 2.6 2 0.8 0003-4975/91/$3.50
HENDREN ET AL ISCHEMIC MITRAL VALVE REPAIR
Ann Thorac Surg 1991;52:124652
Fig 1. From 1985 through 1989, 1,292 patients underwent surgical procedures on their native mitral valve. Ischemic insufficiency was the cause of the mitral disease in 6.5%.
weeks and 98.0 ? 145.8 weeks in the acute and chronic groups, respectively. Degree of mitral insufficiency was calculated by angiography or echocardiography and ranged from 2+ (moderate) to 4+ (severe) with a mean of 3.23 & 0.66. Left ventricular function was assessed by echocardiography or ventriculography, or both, and categorized as normal, mild, moderate, or severe impairment. Twenty-three patients (35.4%)had mild impairment, 20 (30.8%)had moderate, and 22 (33.8%)had severe. Left ventricular function was more impaired in patients with chronic ischemic insufficiency than in those with acute insufficiency (Fig 2). The mean preoperative New York Heart Association functional classification (NYHA FC) was 3.26 ? 0.69. In patients with acute ischemic mitral insufficiency, 5 (45.0%)were in cardiogenic shock preoperatively. Twelve patients (18.5%) had undergone a previous cardiac operation. Preoperative echocardiography was performed in all patients. Valve dysfunction was classified according to the range of motion of the valve leaflets as described by Carpentier [5]. Valve prolapse was found in 26 (40.0%)of the patients. Valve prolapse resulted from papillary muscle rupture in 8 and papillary muscle infarction with elongation in 18. Restricted leaflet motion occurred in 39 patients (60.0%) and was secondary to regional or global ventricular dilatation. Of the patients with acute ischemic
50
-
ACUTE
Mild
CHRONIC
Fig 2 . Compared with patients with acute mitral insufficiency, those with chronic mitral insufficiency had worse ventricular function.
PROLAPSE
1247
RESTRICTED
Fig 3 . Restricted leaflet motion was secondary to regional or global ventricular dilatation. Left ventricular function was more impaired in patients with this valve classification compared with those with valve prolapse.
mitral regurgitation, 5 (45.0%) had prolapse and 6 (55.0%) had restricted motion. Of the patients with chronic regurgitation, 21 (39.0%)had prolapse and 33 (61.0%)restricted leaflet motion. Left ventricular function was significantly worse in patients with restricted leaflet motion (Fig 3). Color flow Doppler echocardiography was used intraoperatively before repair to evaluate the degree of and cause of insufficiency in all patients. All operations were performed using cardiopulmonary bypass with bicaval cannulation, systemic hypothermia, and cold potassium crystalloid cardioplegia during a single period of aortic cross-clamping. All patients had myocardial revascularization. The mean number of grafts was 2.4 ? 1.0 per patient. After completion of distal anastomoses, valve repair was performed transatrially using techniques described by Carpentier [5]. Multiple techniques were employed to achieve valve competence: leaflet resection (3), chordal shortening (15), papillary muscle reimplantation (8), papillary muscle shortening (3), and annuloplasty (63). All but 2 patients had annuloplasty. Mitral valvuloplasty was combined with simultaneous ventricular aneurysm resection in 3 patients. Three patients in the acute group and 5 in the chronic group had insufficiency resulting from papillary muscle rupture and were repaired by papillary muscle reimplantation. Patients demonstrating restricted leaflet motion had valve repair with annuloplasty alone. Early in the series, annuloplasty was performed with a Carpentier-Edwards ring. Recently, annular remodeling has been performed with posterior annular plication with a strip of glutaraldehyde-treated bovine pericardium. Distention of the left ventricle with iced saline solution was used for initial testing of valve repair competence. Intraoperative Doppler echocardiography was used to assess definitive valve competence in all patients immediately after repair. Follow-up data were acquired on all surviving patients. Univariate analyses of hospital and late survival were performed. The clinical variables in the analysis were as follows:
1248
HENDREN ET AL ISCHEMIC MITRAL VALVE REPAIR
Preoperative Clinical Variables Age Sex Preoperative NYHA FC Duration of symptoms Duration of ischemic mitral regurgitation Diabetes Hypertension
Ann Thorac Surg 1991;52:1246-52
Table 1. Perioperative Morbidity (n = 59) No. of Patients
Cause
3 (5.1%) 5 (8.5%) 3 (5.1%) 2 (3.4%)
Cerebral vascular accident Myocardial infarction Reexploration for bleeding Failed valve repair
lrnmediate Preoperative Condition
Stable Shock Intraaortic balloon pump Blood urea nitrogen (mg/100 mL) Creatinine (mg/100 mL) Rhythm
Preoperative Catheterization and Echocardiographic Variables Number of stenotic vessels (greater than 50%) Angiographic/echocardiographic left ventricular function Leaflet motion Angiographic/echocardiographic quantification of mitral regurgitation Angiographic/echocardiographic wall motion abnormality
Operation-Related Variables Mitral anatomy Leaflet motion classification Number of coronary artery bypass grafts Type of repair performed (chordal procedure, papillary muscle procedure, annuloplasty) Associated operations Reoperation Completeness of revascularization
Immediate Postoperative Variables
Transfusion requirement Perioperative myocardial infarction Cerebral vascular accident Wound complication Blood urea nitrogen (mg/100 mL) Dialysis Respiratory failure Prolonged inotropic support (greater than 48 hours) Reexploration for bleeding
7.7% [not significant]). Hospital death occurred in 1 patient with acute disease secondary to postoperative sepsis. Five hospital deaths occurred in patients with chronic ischemic insufficiency, 2 from postoperative sepsis and 3 from cardiac failure. Preoperative variables associated with increased risk for hospital death were poor preoperative NYHA FC ( p = 0.05) and need for preoperative intraaortic balloon pump ( p 5 0.01). Acute presentation, valve motion classification, and technique of repair were not predictive risk factors. It was not possible to identify significant risk factors using multivariate analyses. Perioperative morbidity in 59 surviving patients is listed in Table 1. Prolonged inotropic support was needed in 25 patients (38.5%). Two patients with chronic insufficiency underwent reoperation during initial hospitalization for failed repair. In 1 patient there was partial dehiscence of a Carpentier-Edwards ring and in a second there was suture dehiscence of a repaired leaflet. Follow-up was complete at a mean of 36.6 t 19.0 months. There were 18 late deaths in 59 hospital survivors resulting in 63% survival at 3 years. Patient survival was 96% for valve prolapse and 48% for restrictive leaflet motion ( p = 0.02). Table 2 lists the causes of late deaths. The single late death in a patient with valve prolapse was noncardiac in origin. Univariate analysis identified the following as predictors of late mortality: incomplete revascularization ( p = 0.01), perioperative stroke ( p = 0.002), perioperative myocardial infarction ( p = 0.01), perioperative renal failure ( p < 0.01), prolonged postoperative respiratory insufficiency ( p = 0.006), preoperative intraaortic balloon pump ( p < 0.001), and restrictive leaflet motion ( p = 0.02). It was not possible to identify signifi-
Late Postoperative Variables
Postoperative functional class Postoperative regurgitation Postoperative left ventricular function
Table 2. Causes of Late Deaths No. of Patients
Cause
For hospital survival, p values were calculated by means of a series of x2 tests of association ( p values 5 0.05 were considered statistically significant). To identify risk factors for late mortality, the variables were tested for Wilcoxon rank sum tests (values 5 0.05 significant). Multivariate analyses were not used because of an insufficient number of overall deaths in the study group.
Results There were six in-hospital deaths (9.2%) (acute, 9.1%; chronic 9.3% [not significant]; prolapse, 11.5%;restrictive,
Patients with leaflet prolapse (n = 1) Gastrointestinal bleeding Patients with restrictive leaflet motion (n Congestive heart failure Sudden death Multiorgan failure Arrythmia Renal failure Sepsis Cancer
1(100%) =
18) 6 (35.3%) 4 (23.5%) 2 (11.8%) 1(5.9%) 2 (11.8%) 1(5.9%) 1(5.9%)
HENDREN ET AL ISCHEMIC MITRAL VALVE REPAIR
Ann Thorac Surg 1991;521246-52
1249
Table 3 . Mitral Value Replacement for lschemic Mitral lnsuficiency Series
K
40-
20 -
---
Gerbode et a1 [7] Pinson et a1 [8] Lytle et a1 [9] Connolly et a1 [lo] Kay et a1 [ll] Rankin et a1 [6]
47.7%
Prolapse n=23 Restricted n=36 I
Total NA
cant risk factors using multivariate analysis. Restricted leaflet motion was associated with high risk for late death in patients with both acute and chronic mitral regurgitation (Fig 4). Of the 41 long-term survivors, 29 (70.7%) are in NYHA FC I, 8 (19.5%) in NYHA FC 11, and 4 (9.8%)in NYHA FC 111. The mean degree of postoperative mitral insufficiency was 0.6 ? 0.8 in survivors. Two patients required reoperation for failed repair. Both reoperations occurred in patients with chronic insufficiency and valve prolapse, one as a result of a dehisced Carpentier-Edwards ring and the other as a result of newly ruptured chordae. All patients undergoing reoperation were treated with valve replacement. There were no thromboembolic or other valve-related events.
Comment As the advantages of mitral valve repair are recognized, valvuloplasty for all forms of mitral valve disease is being increasingly applied [4]. Previous surgical management of ischemic mitral regurgitation has been suboptimal. Moderate and severe mitral insufficiency precipitated by myocardial infarction is infrequent, representing the cause of valve pathology in less than 10% of the mitral valve operations [6]. After rheumatic and degenerative pathology, ischemia is the third most common cause of valve dysfunction in patients undergoing operation on their native mitral valves at The Cleveland Clinic Foundation [4]. The patients in this series undergoing mitral valve repair represent 80% of all patients with ischemic regurgitation treated surgically during the period 1985 to 1989. Nineteen patients received prosthetic valve replacement, primarily reflecting surgeon inexperience with repair techniques or preference. Occasionally, valve repair is not appropriate because of associated valve pathology or other major leaflet abnormalities that are more suitably treated with prosthetic valve replacement. Currently, the majority of patients with ischemic regurgitation are amenable to reconstructive techniques.
No. of Patients
41 28 47
Hospital Mortality
Late Survival
48% 38%
NA 44% at 5 years NA 85% at 5 years 35% at 5 years NA
16
10% 19%
40 42
35% 45%
214
33%
not available.
=
Tables 3 and 4 summarize major series of patients with ischemic mitral regurgitation treated by revascularization with mitral valve replacement and repair, respectively. The hospital mortality of 9.2% compares favorably with previously published data [7-121. A lower operative mortality with mitral valve repair for ischemic disease has been previously documented [ll, 121. A statistical comparison of results between repair and replacement was not possible in this study because of the small number of patients having replacement. When compared with previously published data, the operative mortality for repair is lower. Early survival benefit of valvuloplasty for ischemic insufficiency has previously been most marked in patients with recent myocardial infarction [ll, 121. Early results with valve repair for acute ischemic insufficiency showed a mortality of 28% [12]. In this series, there was only one death among patients undergoing valvuloplasty for acute ischemic insufficiency, and acute disease did not emerge as a risk factor for hospital mortality. This may in part reflect improved preoperative and intraoperative management of the ischemic ventricle. However, when compared with patients with chronic ischemic insufficiency, global left ventricular function was better in patients with recent infarction. In this series, only 16.9% of patients requiring operation had acute ischemia. With improvement in the early management of ischemic regurgitation by reperfusion therapy or balloon angioplasty, fewer patients with acute insufficiency require operation [l]. Moderate to severe ischemic regurgitation has been
Table 4. Mitral Valve Repair for lschernic Mitral lnsufficiency Series Kay et a1 [ll] Rankin et a1 [2] Rankin et a1 [6] Current series Total NA
=
not available.
No. of Patients 101
23 40 65 229
Hospital Mortality
Late Survival
15% 26% 18%
64% at 5 years NA NA 63% at 3 years
9%
17%
1250
HENDREN ET AL ISCHEMIC MITRAL VALVE REPAIR
demonstrated to have an adverse influence on long-term survival [l].However, repair of ischemic mitral dysfunction results in better late survival compared with valve replacement [ll].In this study, there were 18 late postoperative deaths. Even though 63% 3-year survival of patients with repair of ischemic dysfunction is poorer than the 5-year survival of 87% for all patients with mitral valvuloplasty at our institution [4], it is similar to longterm rates reported for valvuloplasty for this entity [ll]. Kay and associates [ll]have compiled the largest series of cases of mitral valve repair for ischemic mitral regurgitation and have demonstrated that left ventricular function is the most important determinant of both early and long-term outcome. In this study, all cardiac-related late deaths occurred in patients with restricted leaflet motion in whom ventricular function was most impaired. No late deaths occurred in patients with mild impairment. Compared with prosthetic valve replacement, better preservation of ventricular function (41 is a major advantage of valvuloplasty and makes the argument for mitral valve repair even stronger. The mechanism of insufficiency with ischemic disease is a source of controversy and confusion. The majority of patients with ischemic regurgitation have papillary muscle dysfunction, more recently termed papillary-annular dysfunction, where regurgitation results from a combination of posterior annular dilatation, papillary muscle elongation, and loss of papillary muscle shortening [6, 11, 131. The posterior annular dilatation is thought to be the dominant mechanism of incompetence. Reports also indicate histologic changes indistinguishable from primary myxomatous degeneration may develop in chordae and valve leaflets secondary to ischemia [14]. Patients with ruptured chordae have not been included in this study, as ruptured chordae are more likely to be secondary to primary degenerative changes, which may precede the development of ischemic disease. The high incidence of degenerative mitral valve disease and coronary artery disease makes the coexistence of the two processes likely. In this study, only mitral insufficiency resulting from leaflet prolapse or restricted leaflet motion was found. Posterior annular dilatation, in the absence of degenerative disease of the mitral valve, did not occur. Valve prolapse results from papillary muscle infarction with elongation, paresis of the papillary muscle, or papillary muscle rupture. Restriction of leaflet motion occurs with regional or global ventricular dilatation. Recognizing the functional basis for regurgitation in these valves not only dictates the surgical approach, but also provides for a useful predictor of long-term survival. Previous anatomic studies support this classification. Pathologic examination of ischemic ventricles by Bulkley and colleagues [15] failed to demonstrate annular dilatation, with no correlation between annulus size and the presence or absence of regurgitation. Annular dilatation of any importance has been found to occur only if there is concurrent myxomatous disease. Further, mitral annular dilatation to the degree capable of causing mitral insufficiency occurs infrequently because the surface area of the mitral valve leaflets is 2% times the area of the mitral
Ann Thorac Surg 1991;52124&52
orifice. The mitral ring has a sphincterlike action, making the mitral orifice smaller during systole than during diastole. With ischemia, the majority of dilatation occurs in the mid-portion of the ventricle at the level of origin of the papillary muscles, not the annulus [15]. This change in ventricular geometry causes mechanical displacement resulting in restriction of the leaflet excursion. Impairment of ventricular contraction contributes substantially to the mechanism of regurgitation when papillary muscles are found to be fibrotic. Experimental studies suggest that mitral regurgitation does not occur as a consequence of fibrosis involving only the papillary muscles. If one or both papillary muscles are made fibrotic in a dog by injection of formalin or by ligating the base of the muscle, no regurgitation results. However, if the free wall beneath the papillary muscles is made fibrotic at the same time so that ventricular contraction is impaired, mitral regurgitation results secondary to restriction of motion of a portion of the leaflets [16]. Previous reports concerning ischemic mitral valve repair have demonstrated the efficacy of Kay annuloplasty for this entity [6, 11, 121. We have found valve prolapse to be more common and believe that if valve prolapse is present, annuloplasty alone results in unsuccessful repair. A number of techniques can be used for valve repair for ischemic dysfunction. With valve prolapse, standard Carpentier techniques are successful. For restricted leaflet motion, annuloplasty reduces the size of the annulus, resulting in increased leaflet coaptation and eliminating regurgitation. These concepts can be applied when either acute or chronic disease is present. Patients with papillary muscle rupture can be treated with papillary muscle reimplantation by several methods. Some were repaired by suturing the ruptured muscle to adjacent papillary muscle, and others by suturing the papillary muscle remnant to the free wall of the ventricle. Annuloplasty, whether performed with a prosthetic ring or suture, is an integral part of reconstruction of the valves. Early in this experience, annuloplasty with a Carpentier ring was performed. More recently, a posterior annular plication using a strip of glutaraldehyde-treated bovine pericardium has been used. Experimental fixation of the mitral annulus with a ridged annuloplasty ring has impaired left ventricular systolic function [17]. Further, a rigid ring may interfere with the saddle shape of the mitral annulus, which helps distribute the stress on the mitral valve during systole. Use of a strip of bovine pericardium resulting in multiple small posterior plications may preserve the natural shape of the annulus, and preliminary evidence supports better preservation of ventricular function by this modality. Two in-hospital patients had failed repairs early in the experience before the routine use of transesophageal echocardiography to monitor results. Two additional failures occurred late postoperatively, reflecting the learning curve with this entity. Transventricular repair of ischemic dysfunctions has been advocated [6]; however, we have not adapted this approach. Using cardiac mobilization and a self-retaining retractor, exposure of the mitral valve transatrially is
Ann Thorac Surg 1991;52:1246-52
excellent [18]. The ability to judge the extent of leaflet prolapse is facilitated by a transatrial approach, as is the ability to perform annuloplasty, which is an important component of valve repair for ischemic dysfunction. Ischemic mitral insufficiency is an uncommon cause of mitral valve disease which is amenable to repair in the majority of both acute and chronic cases. Valvuloplasty can be performed with a low operative mortality and is associated with superior late survival in patients with valve prolapse.
References 1. Hickey MS, Smith R, Muhlbaier LH, et al. Current prognosis of ischemic mitral regurgitation. Circulation 1988;78(Suppl 1):51-9. 2. Rankin JS, Hickey MS, Smith LR, et al. Current management of mitral valve incompetence associated with coronary artery disease. J Cardiac Surg 1989;4:2542. 3. Rankin JS, Hickey MS, Smith LR, et al. Ischemic mitral regurgitation. Circulation 1989;79(Suppl 1):116-21. 4. Cosgrove DM, Stewart WJ. Mitral valvuloplasty. Curr Probl Cardiol 1989;14:353-416. 5. Carpentier A. Cardiac valve surgery-the ”French correction.’’ J Thorac Cardiovasc Surg 1983;86:323-37. 6. Rankin JS, Livesey SA, Smith LR, et al. Trends in the surgical treatment of ischemic mitral regurgitation: effects of mitral valve repair on hospital mortality. Semin Thorac Cardiovasc Surg 1989;l:14943. 7. Gerbode FL, Hetzer R, Krebber HJ. Surgical management of papillary muscle rupture due to myocardial infarction. World J Surg 1978;2:791-6.
HENDREN ET AL ISCHEMIC MITRAL VALVE REPAIR
1251
8. Pinson CW, Cobanoglu A, Metzdorff MT, Grunkemeier GL, Kay PH, Starr A. Late surgical results for ischemic mitral regurgitation. J Thorac Cardiovasc Surg 1984;88:66>72. 9. Lytle BW, Cosgrove DM, Gill CC, et al. Mitral valve replacement combined with myocardial revascularization: early and late results for 300 patients, 1970 to 1983. Circulation 1985; 71:1179-90. 10. Connolly MW, Gelbfish JS, Jacobowitz IJ, et al. Surgical results for mitral regurgitation from coronary artery disease. J Thorac Cardiovasc Surg 1986;91:379-88. 11. Kay GL, Kay JH, Zubiate P, Yokoyama T, Mendez M. Mitral valve repair for mitral regurgitation secondary to coronary artery disease. Circulation 1986;74(Suppl 1):8%98. 12. Rankin JS, Feneley MP, Hickey MS, et al. A clinical comparison of mitral valve repair versus valve replacement in ischemic mitral regurgitation. J Thorac Cardiovasc Surg 1988; 95:165-77. 13. Burch GE, DePasquale NP, Phillips JH. The syndrome of papillary muscle dysfunction. Am Heart J 1968;75:399-415. 14. Fishbein MC. Mitral insufficiency in coronary artery disease. Semin Thorac Cardiovasc Surg 1989;1:129-32. 15. Bulkley BH, Roberts WC. Dilatation of the mitral annulus. A rare cause of mitral regurgitation. Am J Med 1975;59:45743. 16. Tsakiris AG, Rastelli GC, Amorim D, Titus JL, Wood EH. Effect of experimental papillary muscle damage on mitral valve closure in intact anesthetized dogs. Mayo Clin Proc 1970;45:27585. 17. David TE, Komeda M, Pollick C, Burns RJ. Mitral valve annuloplasty: the effect of type on left ventricular function. Ann Thorac Surg 1989;47:524-8. 18. Cosgrove DM. A self-retaining retractor for mitral valve operations. J Thorac Cardiovasc Surg 1986;92:305-6.
DISCUSSION D R GREGORY L. KAY (Los Angeles, CA): I would like to thank The Society and the committee for the privilege of opening the discussion of this excellent paper. The authors classify patients with ischemic mitral regurgitation as either having restricted or prolapsing leaflet motion and have demonstrated the apparent impact of this division on survival. Our interest in this disease entity began in the early 1960s, and we have a series of more than 260 patients at this time who have undergone operation for ischemic mitral regurgitation. We have found that the important prognostic determinants for survival of these patients are ejection fraction and New York Heart Association classification. Compromised ventricular function (ejection fraction < 0.4) and New York Heart Association class IV were indicative of poor operative and long-term outcome. Our series also demonstrates that valve repair is superior to valve replacement, particularly for the more compromised patients. Our long-standing interest in mitral valve repair allows us to report on its excellent durability. More than 90% of patients who are alive at 12 years have not required further operation for mitral valve disease. In summary, mitral valve repair is durable and has the most impact on survival in patients who are most compromised. Noel Mills has often stated that, in regard to coronary artery disease, “Anatomy is destiny.” The same statement can be made about this variant, ischemic mitral regurgitation, and as always, the surgeon‘s choice of procedures has a major impact on the immediate and long-term outcome.
I would like to ask Dr Hendren and associates two questions: As you have gained experience with mitral repair, are you now more aggressive in repairing these valves than before? Also, has your technique evolved over time? D R JAMES H. OURY (Missoula, MT): I would like to thank Dr Hendren and Dr Cosgrove for bringing to our attention a very timely subject, namely, the surgical treatment of ischemic mitral valve disease. They have contributed a great deal to this effort in the past, and I am sure they will continue, as we have seen today. My colleagues and I would also make a plea for a classification and a systematic approach to this still rather confusing entity. I would like to make two points and pose two questions for Dr Hendren. In reviewing a consecutive series of 161 patients with combined coronary artery disease and mitral insufficiency over a 5-year period from 1984 to 1989, we found three groups with differing valve pathology, operative mortality, and prognosis. The average age in this group of patients was, again, very similar to Dr Hendren’s series, 67 years of age; the mean follow-up is 2.7 years. The classification fell into three major categories: the first, coronary artery disease with annular dilatation, which would correspond to Dr Hendren’s restricted leaflet motion, we would call a functional classification; the second would be organic, that is, coronary artery disease with ischemic, dysfunctional or infarcted papillary muscle; the third, associated coronary disease with leaflet or chordal pathology. The incidence of repair versus
1252
HENDREN ET AL ISCHEMIC MITRAL VALVE REPAIR
replacement is quite different in these three categories, and the operative mortality is correspondingly different. For instance, in the functional category, coronary artery disease with annular dilatation, 29 of the 32 patients had repair; the operative mortality was 12%, and the actuarial 5-year survival was about the same as in the series presented by Dr Hendren. The very high risk group, which we have classified as the organic group, would be coronary artery disease with ischemic, dysfunctional, or infarcted papillary muscle. The operative mortality is quite a bit higher; however, actuarial survival is about the same. The third group, coronary artery disease with leaflet or chordal pathology, including redo procedures, accounted for the largest number in this particular series, and about half of these patients underwent repair. The operative mortality, as you would expect in this series, is quite low (6%), and the late actuarial 5-year survival is higher. Based on this series we would add two notes of caution. Given this age group, the average age being 69 years, we found that there was no significant difference in intermediate survival whether patients had repair or replacement. Again, I would stress that in the replacement category, we used chordal sparing techniques both in the anterior and posterior leaflet, and I think this is very important in this particular entity. Mitral valve repair, therefore, we thought should be reserved, in general, for younger patients in classifications I and 111. We agree with Dr Hendren’s comment regarding the use of a flexible ring as being important. We favor the flexible ring in all of our annuloplasties. One further point that I would like to introduce is that, in the ongoing debate of flexibility versus the rigid ring annuloplasty, the flexible ring and its recipient do appear to maintain their flexibility at extremes of both altitude and temperature! One of the patients in our series recently trekked to 18,000 feet on Mt Everest. I have two questions for Dr Hendren. Mitral regurgitation in the setting of the ischemic ventricle may be highly variable depending on variations in preload and afterload. Have you devised or used a method of predicting intraoperatively the need for mitral valve repair in this group? Second, did your intermediate-term results demonstrate a statistically significant difference in mortality between mitral valve repair and mitral valve replacement? DR ROBERT W. M. FRATER (Bronx, NY): Thank you for the privilege of discussing this report. It makes an extremely important point. The normal mitral valve, when it opens, moves down into the ventricle as the posterior left ventricular wall elongates. The normal mitral valve closes by annular contraction and also by the movement of the papillary muscles and the cusps up back into the annulus. Patients with posterior infarcts who have grade I11 or IV mitral insufficiency in the coronary care unit needing acute operation most often have failure of shortening of the posterior left ventricular wall as their mechanism of insufficiency. It stays down there during contraction and the cusps do not meet.
Ann Thorac Surg 1991;52124652
You can compensate for that deficiency in posterior left ventricular wall function by performing an annuloplasty; you will notice that in this report, annuloplasty was a fundamental part of the repair and the only repair in a substantial number of cases. 1 think the recognition that annuloplasty will compensate for a failure of shortening of the posterior left ventricular wall transforms operation for acute mitral insufficiency and coronary disease in patients coming out of the critical care unit. We have followed this principle now for some years, and it has made an enormous difference to our approach. It can be seen on echocardiography; you can see the failure of contraction. You can anticipate that with normal cusps and chordae, gross mitral insufficiency can be corrected by a simple annuloplasty. I congratulate Dr Hendren and associates for making this very important point. DR HENDREN:We would like to thank the discussants for their comments. Dr Kay asked whether any changes in our technique have evolved based on our experience. We have been concerned somewhat with the use of a rigid ring, knowing that experimental fixation of the annulus with a rigid ring has impaired systolic function, and, further, that it may interfere with the saddle shape of the annulus, which helps distribute the stress during systole. In the latter part of our experience, we have used a strip of glutaraldehyde-treated bovine pericardium in lieu of a rigid ring, which essentially results in multiple small plications of the posterior annulus. We think that this may help preserve the natural shape of the annulus, and we have preliminary data that suggest the left ventricular function in the short term is better preserved. The issue of whether we are encouraged and are more aggressive with repairing ischemic mitral valves as opposed to valve replacement was raised. The answer is yes. There are certainly rare instances of severe rheumatic disease or severe degenerative disease that may make repair not feasible. However, we believe that because a major advantage of mitral valvuloplasty, when compared with mitral valve replacement, is better preservation of left ventricular function, the more seriously left ventricular function is impaired the more compelling the argument is to proceed with valvuloplasty. Dr Oury asked a question regarding our experience with valve replacement in the setting of ischemic mitral regurgitation. We did not have a large enough number of patients with valve replacement during this time period to compare, and so those data are not available. Our approach to the ischemic valve is that if there is intermittent ischemia, then a mitral valve procedure is not needed. If there is serious regurgitation that cannot be eliminated with medical management, then it is our belief that revascularization alone will not be adequate, and we will proceed with a valve procedure. Once we reach that decision-making point, the decision to proceed with a valve procedure is made as it is for every other cause of mitral insufficiency. Dr Frater, we agree with your assessment of the basis of regurgitation.
(lo),
Over a 5-year period, 1,292 patients had operation on their native mitral valves. Ischemia was the cause of mitral insufficiency in 84 patients (6.5%). Sixty-five patients (77.4%)had mitral valve repair. Mean age was 66 & 10 years; 35 patients (53.8%)were women. Mean degree of preoperative insufficiency was 3.2 & 0.7; mean preoperative New York Heart Association functional class was 3.3 & 0.7. Eleven patients (16.9%)had acute and 54 (83.1%)had chronic mitral insufficiency. Valve prolapse was present in 26 patients (40%). Restrictive leaflet motion secondary to regional or global left ventricular dilatation occurred in 39 patients (60%).All patients had associated myocardial revascularization followed by transatrial valvuloplasty. Multiple techniques were employed to achieve valve competence: leaflet resection (31, chordal shortening (15), papillary muscle reimplantation
papillary muscle shortening (3), and annuloplasty (63). There were six (9.2%)hospital deaths (acute, 9.1%; chronic, 9.3% [not significant]; prolapse, 11.5%; restrictive, 7.7% [not significant]). The mean degree of postoperative mitral insufficiency was 0.6 2 0.8 in 51 patients. At a mean follow-up of 3.1 f 1.6 years, patient survival was 96% for patients with valve prolapse and 48% for those with restrictive leaflet motion ( p = 0.02). New York Heart Association functional class was improved in all groups. Ischemic mitral insufficiency is an uncommon cause of mitral valve disease that is amenable to repair in the majority of cases of both acute and chronic onset. The operative mortality is low, and operation is associated with superior survival in patients with valve prolapse.
I
Patients and Methods
schemic mitral regurgitation remains one of the most challenging management problems in cardiac surgery and is associated with suboptimal results. The majority of patients with ischemic regurgitation are treated medically, but patients with severe regurgitation were found to benefit from surgical treatment [ 1-31. However, the operative morbidity and mortality associated with operation for ischemic mitral regurgitation are higher than those for other forms of mitral valve insufficiency. In the last decade, the operative approach to moderate or severe ischemic regurgitation has evolved from myocardial revascularization to surgical revascularization combined with mitral valvuloplasty. Recognized advantages of mitral valve repair compared with prosthetic valve replacement are lower operative mortality, better preservation of ventricular function, reduced valverelated complications, decreased requirement for anticoagulation therapy, improved long-term survival, and decreased cost [4]. To evaluate the efficacy of mitral valve repair for ischemic regurgitation, our experience over a 4X-year period was reviewed. Presented at the Twenty-seventh Annual Meeting of The Society of Thoracic Surgeons, San Francisco, CA, Feb 1E-20, 1991. Address reprint requests to Dr Cosgrove, Department of Thoracic and Cardiovascular Surgery, The Cleveland Clinic Foundation, 9500 Euclid Ave, Cleveland, OH 44195-5066.
0 1991 by The Society of Thoracic Surgeons
(Ann Thoruc Surg 1991;52:1246-52)
All patients undergoing operation for moderate to severe ischemic mitral insufficiency were included. Ischemic mitral regurgitation was defined as valve insufficiency secondary to myocardial infarction with angiographic or echocardiographic documentation of left ventricular wall motion abnormality in a region perfused by a critically obstructed coronary artery. Patients were excluded if there was any macroscopic or histologic evidence of primary mitral valve pathology. Ischemic regurgitation was considered acute if the interval between myocardial infarction and operation was less than 4 weeks and chronic if the interval was greater. Between January 1985 and June 1989, 1,292 patients underwent valve procedures on their native mitral valves at The Cleveland Clinic Foundation. Eighty-four patients (6.5%)had ischemic mitral insufficiency (Fig 1).Nineteen patients (22.6%) underwent revascularization and valve replacement; 65 (77.4%) underwent revascularization combined with mitral valve repair and form the basis of this report. Eleven patients (16.9%) had acute ischemic insufficiency and 54 (83.1%) had chronic. There were 30 men and 35 women. Patients ranged in age from 47 to 78 years (mean age, 66.0 2 8.1 years). The mean interval from myocardial infarction to operation was 77.1 2 134.6 weeks for the entire group and 2.6 2 0.8 0003-4975/91/$3.50
HENDREN ET AL ISCHEMIC MITRAL VALVE REPAIR
Ann Thorac Surg 1991;52:124652
Fig 1. From 1985 through 1989, 1,292 patients underwent surgical procedures on their native mitral valve. Ischemic insufficiency was the cause of the mitral disease in 6.5%.
weeks and 98.0 ? 145.8 weeks in the acute and chronic groups, respectively. Degree of mitral insufficiency was calculated by angiography or echocardiography and ranged from 2+ (moderate) to 4+ (severe) with a mean of 3.23 & 0.66. Left ventricular function was assessed by echocardiography or ventriculography, or both, and categorized as normal, mild, moderate, or severe impairment. Twenty-three patients (35.4%)had mild impairment, 20 (30.8%)had moderate, and 22 (33.8%)had severe. Left ventricular function was more impaired in patients with chronic ischemic insufficiency than in those with acute insufficiency (Fig 2). The mean preoperative New York Heart Association functional classification (NYHA FC) was 3.26 ? 0.69. In patients with acute ischemic mitral insufficiency, 5 (45.0%)were in cardiogenic shock preoperatively. Twelve patients (18.5%) had undergone a previous cardiac operation. Preoperative echocardiography was performed in all patients. Valve dysfunction was classified according to the range of motion of the valve leaflets as described by Carpentier [5]. Valve prolapse was found in 26 (40.0%)of the patients. Valve prolapse resulted from papillary muscle rupture in 8 and papillary muscle infarction with elongation in 18. Restricted leaflet motion occurred in 39 patients (60.0%) and was secondary to regional or global ventricular dilatation. Of the patients with acute ischemic
50
-
ACUTE
Mild
CHRONIC
Fig 2 . Compared with patients with acute mitral insufficiency, those with chronic mitral insufficiency had worse ventricular function.
PROLAPSE
1247
RESTRICTED
Fig 3 . Restricted leaflet motion was secondary to regional or global ventricular dilatation. Left ventricular function was more impaired in patients with this valve classification compared with those with valve prolapse.
mitral regurgitation, 5 (45.0%) had prolapse and 6 (55.0%) had restricted motion. Of the patients with chronic regurgitation, 21 (39.0%)had prolapse and 33 (61.0%)restricted leaflet motion. Left ventricular function was significantly worse in patients with restricted leaflet motion (Fig 3). Color flow Doppler echocardiography was used intraoperatively before repair to evaluate the degree of and cause of insufficiency in all patients. All operations were performed using cardiopulmonary bypass with bicaval cannulation, systemic hypothermia, and cold potassium crystalloid cardioplegia during a single period of aortic cross-clamping. All patients had myocardial revascularization. The mean number of grafts was 2.4 ? 1.0 per patient. After completion of distal anastomoses, valve repair was performed transatrially using techniques described by Carpentier [5]. Multiple techniques were employed to achieve valve competence: leaflet resection (3), chordal shortening (15), papillary muscle reimplantation (8), papillary muscle shortening (3), and annuloplasty (63). All but 2 patients had annuloplasty. Mitral valvuloplasty was combined with simultaneous ventricular aneurysm resection in 3 patients. Three patients in the acute group and 5 in the chronic group had insufficiency resulting from papillary muscle rupture and were repaired by papillary muscle reimplantation. Patients demonstrating restricted leaflet motion had valve repair with annuloplasty alone. Early in the series, annuloplasty was performed with a Carpentier-Edwards ring. Recently, annular remodeling has been performed with posterior annular plication with a strip of glutaraldehyde-treated bovine pericardium. Distention of the left ventricle with iced saline solution was used for initial testing of valve repair competence. Intraoperative Doppler echocardiography was used to assess definitive valve competence in all patients immediately after repair. Follow-up data were acquired on all surviving patients. Univariate analyses of hospital and late survival were performed. The clinical variables in the analysis were as follows:
1248
HENDREN ET AL ISCHEMIC MITRAL VALVE REPAIR
Preoperative Clinical Variables Age Sex Preoperative NYHA FC Duration of symptoms Duration of ischemic mitral regurgitation Diabetes Hypertension
Ann Thorac Surg 1991;52:1246-52
Table 1. Perioperative Morbidity (n = 59) No. of Patients
Cause
3 (5.1%) 5 (8.5%) 3 (5.1%) 2 (3.4%)
Cerebral vascular accident Myocardial infarction Reexploration for bleeding Failed valve repair
lrnmediate Preoperative Condition
Stable Shock Intraaortic balloon pump Blood urea nitrogen (mg/100 mL) Creatinine (mg/100 mL) Rhythm
Preoperative Catheterization and Echocardiographic Variables Number of stenotic vessels (greater than 50%) Angiographic/echocardiographic left ventricular function Leaflet motion Angiographic/echocardiographic quantification of mitral regurgitation Angiographic/echocardiographic wall motion abnormality
Operation-Related Variables Mitral anatomy Leaflet motion classification Number of coronary artery bypass grafts Type of repair performed (chordal procedure, papillary muscle procedure, annuloplasty) Associated operations Reoperation Completeness of revascularization
Immediate Postoperative Variables
Transfusion requirement Perioperative myocardial infarction Cerebral vascular accident Wound complication Blood urea nitrogen (mg/100 mL) Dialysis Respiratory failure Prolonged inotropic support (greater than 48 hours) Reexploration for bleeding
7.7% [not significant]). Hospital death occurred in 1 patient with acute disease secondary to postoperative sepsis. Five hospital deaths occurred in patients with chronic ischemic insufficiency, 2 from postoperative sepsis and 3 from cardiac failure. Preoperative variables associated with increased risk for hospital death were poor preoperative NYHA FC ( p = 0.05) and need for preoperative intraaortic balloon pump ( p 5 0.01). Acute presentation, valve motion classification, and technique of repair were not predictive risk factors. It was not possible to identify significant risk factors using multivariate analyses. Perioperative morbidity in 59 surviving patients is listed in Table 1. Prolonged inotropic support was needed in 25 patients (38.5%). Two patients with chronic insufficiency underwent reoperation during initial hospitalization for failed repair. In 1 patient there was partial dehiscence of a Carpentier-Edwards ring and in a second there was suture dehiscence of a repaired leaflet. Follow-up was complete at a mean of 36.6 t 19.0 months. There were 18 late deaths in 59 hospital survivors resulting in 63% survival at 3 years. Patient survival was 96% for valve prolapse and 48% for restrictive leaflet motion ( p = 0.02). Table 2 lists the causes of late deaths. The single late death in a patient with valve prolapse was noncardiac in origin. Univariate analysis identified the following as predictors of late mortality: incomplete revascularization ( p = 0.01), perioperative stroke ( p = 0.002), perioperative myocardial infarction ( p = 0.01), perioperative renal failure ( p < 0.01), prolonged postoperative respiratory insufficiency ( p = 0.006), preoperative intraaortic balloon pump ( p < 0.001), and restrictive leaflet motion ( p = 0.02). It was not possible to identify signifi-
Late Postoperative Variables
Postoperative functional class Postoperative regurgitation Postoperative left ventricular function
Table 2. Causes of Late Deaths No. of Patients
Cause
For hospital survival, p values were calculated by means of a series of x2 tests of association ( p values 5 0.05 were considered statistically significant). To identify risk factors for late mortality, the variables were tested for Wilcoxon rank sum tests (values 5 0.05 significant). Multivariate analyses were not used because of an insufficient number of overall deaths in the study group.
Results There were six in-hospital deaths (9.2%) (acute, 9.1%; chronic 9.3% [not significant]; prolapse, 11.5%;restrictive,
Patients with leaflet prolapse (n = 1) Gastrointestinal bleeding Patients with restrictive leaflet motion (n Congestive heart failure Sudden death Multiorgan failure Arrythmia Renal failure Sepsis Cancer
1(100%) =
18) 6 (35.3%) 4 (23.5%) 2 (11.8%) 1(5.9%) 2 (11.8%) 1(5.9%) 1(5.9%)
HENDREN ET AL ISCHEMIC MITRAL VALVE REPAIR
Ann Thorac Surg 1991;521246-52
1249
Table 3 . Mitral Value Replacement for lschemic Mitral lnsuficiency Series
K
40-
20 -
---
Gerbode et a1 [7] Pinson et a1 [8] Lytle et a1 [9] Connolly et a1 [lo] Kay et a1 [ll] Rankin et a1 [6]
47.7%
Prolapse n=23 Restricted n=36 I
Total NA
cant risk factors using multivariate analysis. Restricted leaflet motion was associated with high risk for late death in patients with both acute and chronic mitral regurgitation (Fig 4). Of the 41 long-term survivors, 29 (70.7%) are in NYHA FC I, 8 (19.5%) in NYHA FC 11, and 4 (9.8%)in NYHA FC 111. The mean degree of postoperative mitral insufficiency was 0.6 ? 0.8 in survivors. Two patients required reoperation for failed repair. Both reoperations occurred in patients with chronic insufficiency and valve prolapse, one as a result of a dehisced Carpentier-Edwards ring and the other as a result of newly ruptured chordae. All patients undergoing reoperation were treated with valve replacement. There were no thromboembolic or other valve-related events.
Comment As the advantages of mitral valve repair are recognized, valvuloplasty for all forms of mitral valve disease is being increasingly applied [4]. Previous surgical management of ischemic mitral regurgitation has been suboptimal. Moderate and severe mitral insufficiency precipitated by myocardial infarction is infrequent, representing the cause of valve pathology in less than 10% of the mitral valve operations [6]. After rheumatic and degenerative pathology, ischemia is the third most common cause of valve dysfunction in patients undergoing operation on their native mitral valves at The Cleveland Clinic Foundation [4]. The patients in this series undergoing mitral valve repair represent 80% of all patients with ischemic regurgitation treated surgically during the period 1985 to 1989. Nineteen patients received prosthetic valve replacement, primarily reflecting surgeon inexperience with repair techniques or preference. Occasionally, valve repair is not appropriate because of associated valve pathology or other major leaflet abnormalities that are more suitably treated with prosthetic valve replacement. Currently, the majority of patients with ischemic regurgitation are amenable to reconstructive techniques.
No. of Patients
41 28 47
Hospital Mortality
Late Survival
48% 38%
NA 44% at 5 years NA 85% at 5 years 35% at 5 years NA
16
10% 19%
40 42
35% 45%
214
33%
not available.
=
Tables 3 and 4 summarize major series of patients with ischemic mitral regurgitation treated by revascularization with mitral valve replacement and repair, respectively. The hospital mortality of 9.2% compares favorably with previously published data [7-121. A lower operative mortality with mitral valve repair for ischemic disease has been previously documented [ll, 121. A statistical comparison of results between repair and replacement was not possible in this study because of the small number of patients having replacement. When compared with previously published data, the operative mortality for repair is lower. Early survival benefit of valvuloplasty for ischemic insufficiency has previously been most marked in patients with recent myocardial infarction [ll, 121. Early results with valve repair for acute ischemic insufficiency showed a mortality of 28% [12]. In this series, there was only one death among patients undergoing valvuloplasty for acute ischemic insufficiency, and acute disease did not emerge as a risk factor for hospital mortality. This may in part reflect improved preoperative and intraoperative management of the ischemic ventricle. However, when compared with patients with chronic ischemic insufficiency, global left ventricular function was better in patients with recent infarction. In this series, only 16.9% of patients requiring operation had acute ischemia. With improvement in the early management of ischemic regurgitation by reperfusion therapy or balloon angioplasty, fewer patients with acute insufficiency require operation [l]. Moderate to severe ischemic regurgitation has been
Table 4. Mitral Valve Repair for lschernic Mitral lnsufficiency Series Kay et a1 [ll] Rankin et a1 [2] Rankin et a1 [6] Current series Total NA
=
not available.
No. of Patients 101
23 40 65 229
Hospital Mortality
Late Survival
15% 26% 18%
64% at 5 years NA NA 63% at 3 years
9%
17%
1250
HENDREN ET AL ISCHEMIC MITRAL VALVE REPAIR
demonstrated to have an adverse influence on long-term survival [l].However, repair of ischemic mitral dysfunction results in better late survival compared with valve replacement [ll].In this study, there were 18 late postoperative deaths. Even though 63% 3-year survival of patients with repair of ischemic dysfunction is poorer than the 5-year survival of 87% for all patients with mitral valvuloplasty at our institution [4], it is similar to longterm rates reported for valvuloplasty for this entity [ll]. Kay and associates [ll]have compiled the largest series of cases of mitral valve repair for ischemic mitral regurgitation and have demonstrated that left ventricular function is the most important determinant of both early and long-term outcome. In this study, all cardiac-related late deaths occurred in patients with restricted leaflet motion in whom ventricular function was most impaired. No late deaths occurred in patients with mild impairment. Compared with prosthetic valve replacement, better preservation of ventricular function (41 is a major advantage of valvuloplasty and makes the argument for mitral valve repair even stronger. The mechanism of insufficiency with ischemic disease is a source of controversy and confusion. The majority of patients with ischemic regurgitation have papillary muscle dysfunction, more recently termed papillary-annular dysfunction, where regurgitation results from a combination of posterior annular dilatation, papillary muscle elongation, and loss of papillary muscle shortening [6, 11, 131. The posterior annular dilatation is thought to be the dominant mechanism of incompetence. Reports also indicate histologic changes indistinguishable from primary myxomatous degeneration may develop in chordae and valve leaflets secondary to ischemia [14]. Patients with ruptured chordae have not been included in this study, as ruptured chordae are more likely to be secondary to primary degenerative changes, which may precede the development of ischemic disease. The high incidence of degenerative mitral valve disease and coronary artery disease makes the coexistence of the two processes likely. In this study, only mitral insufficiency resulting from leaflet prolapse or restricted leaflet motion was found. Posterior annular dilatation, in the absence of degenerative disease of the mitral valve, did not occur. Valve prolapse results from papillary muscle infarction with elongation, paresis of the papillary muscle, or papillary muscle rupture. Restriction of leaflet motion occurs with regional or global ventricular dilatation. Recognizing the functional basis for regurgitation in these valves not only dictates the surgical approach, but also provides for a useful predictor of long-term survival. Previous anatomic studies support this classification. Pathologic examination of ischemic ventricles by Bulkley and colleagues [15] failed to demonstrate annular dilatation, with no correlation between annulus size and the presence or absence of regurgitation. Annular dilatation of any importance has been found to occur only if there is concurrent myxomatous disease. Further, mitral annular dilatation to the degree capable of causing mitral insufficiency occurs infrequently because the surface area of the mitral valve leaflets is 2% times the area of the mitral
Ann Thorac Surg 1991;52124&52
orifice. The mitral ring has a sphincterlike action, making the mitral orifice smaller during systole than during diastole. With ischemia, the majority of dilatation occurs in the mid-portion of the ventricle at the level of origin of the papillary muscles, not the annulus [15]. This change in ventricular geometry causes mechanical displacement resulting in restriction of the leaflet excursion. Impairment of ventricular contraction contributes substantially to the mechanism of regurgitation when papillary muscles are found to be fibrotic. Experimental studies suggest that mitral regurgitation does not occur as a consequence of fibrosis involving only the papillary muscles. If one or both papillary muscles are made fibrotic in a dog by injection of formalin or by ligating the base of the muscle, no regurgitation results. However, if the free wall beneath the papillary muscles is made fibrotic at the same time so that ventricular contraction is impaired, mitral regurgitation results secondary to restriction of motion of a portion of the leaflets [16]. Previous reports concerning ischemic mitral valve repair have demonstrated the efficacy of Kay annuloplasty for this entity [6, 11, 121. We have found valve prolapse to be more common and believe that if valve prolapse is present, annuloplasty alone results in unsuccessful repair. A number of techniques can be used for valve repair for ischemic dysfunction. With valve prolapse, standard Carpentier techniques are successful. For restricted leaflet motion, annuloplasty reduces the size of the annulus, resulting in increased leaflet coaptation and eliminating regurgitation. These concepts can be applied when either acute or chronic disease is present. Patients with papillary muscle rupture can be treated with papillary muscle reimplantation by several methods. Some were repaired by suturing the ruptured muscle to adjacent papillary muscle, and others by suturing the papillary muscle remnant to the free wall of the ventricle. Annuloplasty, whether performed with a prosthetic ring or suture, is an integral part of reconstruction of the valves. Early in this experience, annuloplasty with a Carpentier ring was performed. More recently, a posterior annular plication using a strip of glutaraldehyde-treated bovine pericardium has been used. Experimental fixation of the mitral annulus with a ridged annuloplasty ring has impaired left ventricular systolic function [17]. Further, a rigid ring may interfere with the saddle shape of the mitral annulus, which helps distribute the stress on the mitral valve during systole. Use of a strip of bovine pericardium resulting in multiple small posterior plications may preserve the natural shape of the annulus, and preliminary evidence supports better preservation of ventricular function by this modality. Two in-hospital patients had failed repairs early in the experience before the routine use of transesophageal echocardiography to monitor results. Two additional failures occurred late postoperatively, reflecting the learning curve with this entity. Transventricular repair of ischemic dysfunctions has been advocated [6]; however, we have not adapted this approach. Using cardiac mobilization and a self-retaining retractor, exposure of the mitral valve transatrially is
Ann Thorac Surg 1991;52:1246-52
excellent [18]. The ability to judge the extent of leaflet prolapse is facilitated by a transatrial approach, as is the ability to perform annuloplasty, which is an important component of valve repair for ischemic dysfunction. Ischemic mitral insufficiency is an uncommon cause of mitral valve disease which is amenable to repair in the majority of both acute and chronic cases. Valvuloplasty can be performed with a low operative mortality and is associated with superior late survival in patients with valve prolapse.
References 1. Hickey MS, Smith R, Muhlbaier LH, et al. Current prognosis of ischemic mitral regurgitation. Circulation 1988;78(Suppl 1):51-9. 2. Rankin JS, Hickey MS, Smith LR, et al. Current management of mitral valve incompetence associated with coronary artery disease. J Cardiac Surg 1989;4:2542. 3. Rankin JS, Hickey MS, Smith LR, et al. Ischemic mitral regurgitation. Circulation 1989;79(Suppl 1):116-21. 4. Cosgrove DM, Stewart WJ. Mitral valvuloplasty. Curr Probl Cardiol 1989;14:353-416. 5. Carpentier A. Cardiac valve surgery-the ”French correction.’’ J Thorac Cardiovasc Surg 1983;86:323-37. 6. Rankin JS, Livesey SA, Smith LR, et al. Trends in the surgical treatment of ischemic mitral regurgitation: effects of mitral valve repair on hospital mortality. Semin Thorac Cardiovasc Surg 1989;l:14943. 7. Gerbode FL, Hetzer R, Krebber HJ. Surgical management of papillary muscle rupture due to myocardial infarction. World J Surg 1978;2:791-6.
HENDREN ET AL ISCHEMIC MITRAL VALVE REPAIR
1251
8. Pinson CW, Cobanoglu A, Metzdorff MT, Grunkemeier GL, Kay PH, Starr A. Late surgical results for ischemic mitral regurgitation. J Thorac Cardiovasc Surg 1984;88:66>72. 9. Lytle BW, Cosgrove DM, Gill CC, et al. Mitral valve replacement combined with myocardial revascularization: early and late results for 300 patients, 1970 to 1983. Circulation 1985; 71:1179-90. 10. Connolly MW, Gelbfish JS, Jacobowitz IJ, et al. Surgical results for mitral regurgitation from coronary artery disease. J Thorac Cardiovasc Surg 1986;91:379-88. 11. Kay GL, Kay JH, Zubiate P, Yokoyama T, Mendez M. Mitral valve repair for mitral regurgitation secondary to coronary artery disease. Circulation 1986;74(Suppl 1):8%98. 12. Rankin JS, Feneley MP, Hickey MS, et al. A clinical comparison of mitral valve repair versus valve replacement in ischemic mitral regurgitation. J Thorac Cardiovasc Surg 1988; 95:165-77. 13. Burch GE, DePasquale NP, Phillips JH. The syndrome of papillary muscle dysfunction. Am Heart J 1968;75:399-415. 14. Fishbein MC. Mitral insufficiency in coronary artery disease. Semin Thorac Cardiovasc Surg 1989;1:129-32. 15. Bulkley BH, Roberts WC. Dilatation of the mitral annulus. A rare cause of mitral regurgitation. Am J Med 1975;59:45743. 16. Tsakiris AG, Rastelli GC, Amorim D, Titus JL, Wood EH. Effect of experimental papillary muscle damage on mitral valve closure in intact anesthetized dogs. Mayo Clin Proc 1970;45:27585. 17. David TE, Komeda M, Pollick C, Burns RJ. Mitral valve annuloplasty: the effect of type on left ventricular function. Ann Thorac Surg 1989;47:524-8. 18. Cosgrove DM. A self-retaining retractor for mitral valve operations. J Thorac Cardiovasc Surg 1986;92:305-6.
DISCUSSION D R GREGORY L. KAY (Los Angeles, CA): I would like to thank The Society and the committee for the privilege of opening the discussion of this excellent paper. The authors classify patients with ischemic mitral regurgitation as either having restricted or prolapsing leaflet motion and have demonstrated the apparent impact of this division on survival. Our interest in this disease entity began in the early 1960s, and we have a series of more than 260 patients at this time who have undergone operation for ischemic mitral regurgitation. We have found that the important prognostic determinants for survival of these patients are ejection fraction and New York Heart Association classification. Compromised ventricular function (ejection fraction < 0.4) and New York Heart Association class IV were indicative of poor operative and long-term outcome. Our series also demonstrates that valve repair is superior to valve replacement, particularly for the more compromised patients. Our long-standing interest in mitral valve repair allows us to report on its excellent durability. More than 90% of patients who are alive at 12 years have not required further operation for mitral valve disease. In summary, mitral valve repair is durable and has the most impact on survival in patients who are most compromised. Noel Mills has often stated that, in regard to coronary artery disease, “Anatomy is destiny.” The same statement can be made about this variant, ischemic mitral regurgitation, and as always, the surgeon‘s choice of procedures has a major impact on the immediate and long-term outcome.
I would like to ask Dr Hendren and associates two questions: As you have gained experience with mitral repair, are you now more aggressive in repairing these valves than before? Also, has your technique evolved over time? D R JAMES H. OURY (Missoula, MT): I would like to thank Dr Hendren and Dr Cosgrove for bringing to our attention a very timely subject, namely, the surgical treatment of ischemic mitral valve disease. They have contributed a great deal to this effort in the past, and I am sure they will continue, as we have seen today. My colleagues and I would also make a plea for a classification and a systematic approach to this still rather confusing entity. I would like to make two points and pose two questions for Dr Hendren. In reviewing a consecutive series of 161 patients with combined coronary artery disease and mitral insufficiency over a 5-year period from 1984 to 1989, we found three groups with differing valve pathology, operative mortality, and prognosis. The average age in this group of patients was, again, very similar to Dr Hendren’s series, 67 years of age; the mean follow-up is 2.7 years. The classification fell into three major categories: the first, coronary artery disease with annular dilatation, which would correspond to Dr Hendren’s restricted leaflet motion, we would call a functional classification; the second would be organic, that is, coronary artery disease with ischemic, dysfunctional or infarcted papillary muscle; the third, associated coronary disease with leaflet or chordal pathology. The incidence of repair versus
1252
HENDREN ET AL ISCHEMIC MITRAL VALVE REPAIR
replacement is quite different in these three categories, and the operative mortality is correspondingly different. For instance, in the functional category, coronary artery disease with annular dilatation, 29 of the 32 patients had repair; the operative mortality was 12%, and the actuarial 5-year survival was about the same as in the series presented by Dr Hendren. The very high risk group, which we have classified as the organic group, would be coronary artery disease with ischemic, dysfunctional, or infarcted papillary muscle. The operative mortality is quite a bit higher; however, actuarial survival is about the same. The third group, coronary artery disease with leaflet or chordal pathology, including redo procedures, accounted for the largest number in this particular series, and about half of these patients underwent repair. The operative mortality, as you would expect in this series, is quite low (6%), and the late actuarial 5-year survival is higher. Based on this series we would add two notes of caution. Given this age group, the average age being 69 years, we found that there was no significant difference in intermediate survival whether patients had repair or replacement. Again, I would stress that in the replacement category, we used chordal sparing techniques both in the anterior and posterior leaflet, and I think this is very important in this particular entity. Mitral valve repair, therefore, we thought should be reserved, in general, for younger patients in classifications I and 111. We agree with Dr Hendren’s comment regarding the use of a flexible ring as being important. We favor the flexible ring in all of our annuloplasties. One further point that I would like to introduce is that, in the ongoing debate of flexibility versus the rigid ring annuloplasty, the flexible ring and its recipient do appear to maintain their flexibility at extremes of both altitude and temperature! One of the patients in our series recently trekked to 18,000 feet on Mt Everest. I have two questions for Dr Hendren. Mitral regurgitation in the setting of the ischemic ventricle may be highly variable depending on variations in preload and afterload. Have you devised or used a method of predicting intraoperatively the need for mitral valve repair in this group? Second, did your intermediate-term results demonstrate a statistically significant difference in mortality between mitral valve repair and mitral valve replacement? DR ROBERT W. M. FRATER (Bronx, NY): Thank you for the privilege of discussing this report. It makes an extremely important point. The normal mitral valve, when it opens, moves down into the ventricle as the posterior left ventricular wall elongates. The normal mitral valve closes by annular contraction and also by the movement of the papillary muscles and the cusps up back into the annulus. Patients with posterior infarcts who have grade I11 or IV mitral insufficiency in the coronary care unit needing acute operation most often have failure of shortening of the posterior left ventricular wall as their mechanism of insufficiency. It stays down there during contraction and the cusps do not meet.
Ann Thorac Surg 1991;52124652
You can compensate for that deficiency in posterior left ventricular wall function by performing an annuloplasty; you will notice that in this report, annuloplasty was a fundamental part of the repair and the only repair in a substantial number of cases. 1 think the recognition that annuloplasty will compensate for a failure of shortening of the posterior left ventricular wall transforms operation for acute mitral insufficiency and coronary disease in patients coming out of the critical care unit. We have followed this principle now for some years, and it has made an enormous difference to our approach. It can be seen on echocardiography; you can see the failure of contraction. You can anticipate that with normal cusps and chordae, gross mitral insufficiency can be corrected by a simple annuloplasty. I congratulate Dr Hendren and associates for making this very important point. DR HENDREN:We would like to thank the discussants for their comments. Dr Kay asked whether any changes in our technique have evolved based on our experience. We have been concerned somewhat with the use of a rigid ring, knowing that experimental fixation of the annulus with a rigid ring has impaired systolic function, and, further, that it may interfere with the saddle shape of the annulus, which helps distribute the stress during systole. In the latter part of our experience, we have used a strip of glutaraldehyde-treated bovine pericardium in lieu of a rigid ring, which essentially results in multiple small plications of the posterior annulus. We think that this may help preserve the natural shape of the annulus, and we have preliminary data that suggest the left ventricular function in the short term is better preserved. The issue of whether we are encouraged and are more aggressive with repairing ischemic mitral valves as opposed to valve replacement was raised. The answer is yes. There are certainly rare instances of severe rheumatic disease or severe degenerative disease that may make repair not feasible. However, we believe that because a major advantage of mitral valvuloplasty, when compared with mitral valve replacement, is better preservation of left ventricular function, the more seriously left ventricular function is impaired the more compelling the argument is to proceed with valvuloplasty. Dr Oury asked a question regarding our experience with valve replacement in the setting of ischemic mitral regurgitation. We did not have a large enough number of patients with valve replacement during this time period to compare, and so those data are not available. Our approach to the ischemic valve is that if there is intermittent ischemia, then a mitral valve procedure is not needed. If there is serious regurgitation that cannot be eliminated with medical management, then it is our belief that revascularization alone will not be adequate, and we will proceed with a valve procedure. Once we reach that decision-making point, the decision to proceed with a valve procedure is made as it is for every other cause of mitral insufficiency. Dr Frater, we agree with your assessment of the basis of regurgitation.