- Email: [email protected]
Reconstructive surgery of mitral valve incompetence
J
THORAC CARDIOVASC SURG
79:338-348, 1980
Reconstructive surgery of mitral valve incompetence Ten-year appraisal Between January, 1969, and January, 1978, 551 patients with mitral incompetence were treated by a system of reconstructive techniques. Mitral valve incompetence was classified into three types according to leaflet pliability; type 1 normal leaflet motion, 150 cases; type Il, leaflet prolapse, 213 cases; and type l1l, restricted leaflet motion, 188 cases. Associated tricuspid valvular disease was present in 174 cases (31.5%) and was treated by prosthetic ring annuloplasty. The operative mortality rate was 4.2% (16/377) in the mitral group and 14% (25/174) in the mitral-tricuspid group. Follow-up data are available in 341 patients from 1 year to 10 years (average 4 112 years). The late mortality rate was 7% (24/341). Actuarial curves including hospital mortality rate show an 82% survival rate at 9 years in the mitral group and a 79% rate in the mitral-tricuspid group. Thirty-seven patients (11%) underwent reoperation mainly for residual (17) or recurrent (16) mitral incompetence. Thromboembolism occurred in 12 patients for an embolic rate of 0.6% per patient-year, even though 48% were not given anticoagulants. A ccording to the New York Heart Association (N.Y.HA .) classification, 76% (207/270) of the patients were in Class I, 19% (51/270) were in Class ll, 4% (10/270) were in Class l1l, and 0.7% were in Class IV (2/270). Results of postoperative catheterization and angiocardiography are available in 52 patients. Comparison between the various groups shows that the best results were obtained in type II mitral incompetence, followed by type 1 and type III mitral incompetence. This experience demonstrates that predictable and stable long-term results have been achieved by techniques of valvular reconstruction with a low incidence of thromboembolism. Reproducibility of the techniques is a limiting factor which can be overcome by adequate training and progressive experience. Patient selection is based on the valvular disease rather than age, physical condition, or cause of valvular disease.
A. Carpentier, M.D. (by invitation), S. Chauvaud, M.D. (by invitation), J. N. Fabiani, M.D. (by invitation), A. Deloche, M.D. (by invitation),
J. ReIland, M.D. (by invitation), A. Lessana, M.D. (by invitation), Cl. d'Allaines, M.D. (by invitation), Ph. Blondeau, M.D. (by invitation), A. Piwnica, M.D. (by invitation), and Ch. Dubost, M.D., Paris, France
Reconstructive surgery of acquired mitral valve incompetence continues to raise three controversial problems: (l) predictability of results, (2) reproducibility of techniques, and (3) selection of patients. Having reviewed our experience with 551 mitral valve reconstructions performed at the Clinique de Chirurgie CarFrom the Chaire de Chirurgie Cardiovasculaire, Hopital Broussais, University of Paris, Paris, France. Supported by a grant from the C.N.R.S. Read at the Fifty-ninth Annual Meeting of the American Association for Thoracic Surgery, Boston, Mass., April 30 to May 2, 1979. Address for reprints: A. Carpentier, M.D., Hopital Broussais, 96 rue Didot, 75014-Paris (France).
338
diovasculaire of the Hopital Broussais in Paris, France, between January, 1969, and January, 1978, we will attempt to answer these points. Excluded from this study were 75 cases of congenital mitral valve disease and 286 cases of mitral valve incompetence associated with aortic valve replacement, which brings the total to 912 mitral valve reconstructions. Also excluded from the study were operations for pure mitral valve stenosis. Patients and methods During the period from January, 1969, to January, 1978, 551 patients with mitral valve incompetence were treated by reconstructive techniques. The ages of
0022-5223/80/030338+11$01.10/0 © 1980 The C. V. Mosby Co.
Volume 79
339
Mitral incompetence
Number 3
March,1980
Table I. Pathophysiological classification and valvular lesions* Normal leaflet motion Anulus dilatation Leaflet perforation Type II: Prolapsed leaflet Ruptured chordae Elongated chordae Ruptured papillary muscle Type III: Restricted leaflet motion Commissural fusion and leaflet thickening Associated fused chordae
Table III. Preoperative clinical condition of the patients
Type I:
159 157 2 213 109 103 I 179 52 127
'Calcifications were present in 26 cases.
Table II. Cause of valve lesions (in 419 cases) Cause
Rheumatic fever Endocarditis Myxoid degeneration (floppy valve) Fibroelastic deficiency Ischemic disease
~ 97 2 4 6
Type II
6I 10 49 30 I
IType III 159
the patients ranged from 2 to 77 years (average 37 years). Sixty patients were younger than 15 years of age. Pathophysiological classification. Mitral valve incompetence was classified into three types according to leaflet pliability!' 2; type I, normal leaflet motion; type II, prolapsed leaflet; and type III, restricted leaflet motion. Leaflet motion was assessed prior to operation by echocardiography or angiocardiography (or both) and during operation by mobilization of the leaflets with a nerve hook and by examination of the leaflets in the beating heart. In type I, the free edges of the leaflets remain below the plane of the anulus during systole and open normally during diastole. In type II, the free edge of one or both leaflets overrides the plane of the anulus during systole. In type III, one or both leaflets do not open fully during diastole. The classification is further defined as to the primary lesion. Table I categorizes the patients with regard to types and lesions. When associated lesions are present, the predominant lesion serves to classify the patient. For example, a patient with severe prolapse of the anterior leaflet owing to elongated chordae and some degree of restricted motion of the mural leaflet would be considered to have type II disease. The causes of the valve lesions were recognized in 419 patients and are listed in Table II. In nonrheumatic mitral valve insufficiency, we found it useful to introduce a clear distinction between fibroelastic deficiency and myxoid degeneration based on clinical patterns, gross structure, and histologic examination
N.Y.H.A. Class II N.Y.H.A. Class III N.Y.H.A. Class IV Acute rheumatic fever Angina Myocardial infarction Systemic embolism Giant atrium Sinus rhythm Atrial fibrillation Previous commissurotomy Previous plication of mitral anulus Moderate aortic insufficiency Tricuspid insufficiency
57 223 271 7
5 3 4 9 115 436 7
2 6 181
Legend: N.Y.H.A., New York Heart Association.
Table IV. Combined procedures used in 198 patients Atrial septal defect repair Coronary bypass graft Reduction of giant left atrium Resection of ventricular aneurysm Tricuspid ring annuloplasty
10 4 8 2
174
of the mitral valve (Fig. 1). Fibroelastic deficiency did not seem to be secondary to coronary artery disease, as the incidence of coronary artery disease in this series was small according to the age of the patients. Clinical condition of the patients. Patients were evaluated on the basis of the New York Heart Association (N.Y.H.A.) classification. The condition prior to operation is outlined in Table III. Patient selection for mitral valve repair was based on neither the clinical condition nor the age of the patient. Acute rheumatic fever associated with severe cardiac failure was not considered to be a contraindication, and six such patients were operated upon while receiving corticosteroid treatment. Indications were based upon the actual valvular lesions: All patients with noncalcified mitral valvular disease except those with acute bacterial endocarditis were considered to be candidates for valve repair. The percentage of noncalcified valves which were repaired varied from 15% to 80% according to the individual surgeon. Twenty-six patients with calcified mitral valvular disease were unexpectedly treated by reconstructive procedures because the lesions were found to be suitable for repair at operation. The duration of symptoms varied between 3 weeks and 18 years. Ten patients had previous commissurotomy and two had previous annuloplasties by the plication technique; combined procedures are listed in Table IV.
340 Carpentier et at.
The Journal of Thoracic and Cardiovascular Surgery
Fig. 1. Characteristic features in myxoid degeneration (Aj and fibroelastic deficiency (B j. Myxoid degeneration: Redundant, thickened, and yellowish leaflets with excess tissue. Severe dilatation of anulus. Elongated and irregular (thin or thickened) chordae. Disruption of collagen and elastic layers, and excess acid mucopolysaccharides. Average age at operation-48 years. Average time interval between first clinical symptom and operation-28 years. Fibroelastic deficiency (pellucid mitral valve): Smooth, thin, and translucent leaflets (no excess tissue). Moderate dilatation of anulus. Slightly elongated and very thin chordae. Fibroelastic and acid mucopolysaccharidedeficiency. Average age at operation-61 years. Average time interval between first clinical symptom and operation-6 years.
Surgical principles The following results are those obtained with a system of valvular reconstruction comprising four basic techniques that were used either individually or in association depending on the valvular lesions (Fig. 2): (l) anulus remodelling with a prosthetic ring, (2) leaflet tissue resection and repair, (3) chordal shortening, and (4) chordal fenestration and resection. Because technical details have been presented in previous papers, 1-5 we will recall here only the basic principles of these techniques and the occasional changes which have been introduced since previous publications. Anulus remodelling. Dilatation of the anulus fibrosus is recognized as one of the major factors in mitral insufficiency. The dilatation has certain characteristic features which must guide the technique of reconstruction: 1. Dilatation affects the mural leaflet and the commissural areas.
2. Dilatation does not affect the attachment of the aortic leaflet. 3. Dilatation is associated with a deformation of the anulus such that the anteroposterior diameter of the orifice is greater than the transverse diameter. 4. Dilatation is an evolving process which should be controlled by stabilizing the entire anulus. Prosthetic rings of a suitable shape and size are necessary to perform a measured annuloplasty which will restore the normal contour and thereby both a normal orificial area and optimum function of the valve. The selective flexibility of the prosthetic ring allows a more physiological repair. Prosthetic ring implantation does not present any surgical difficulty. It takes 20 to 30 minutes, depending on the surgeon's experience, and requires only that the spatial relationships between sutures on the anulus and on the prosthetic ring be carefully matched. Leaflet resection. Resection of abnormal segments
Volume 79
Mitral incompetence
Number 3
341
March . 1980
annulus remodelling
chordae shortening
leaflet
chordae
resect ion-suture
fenestration
Fig. 2. The four basic techniques used in reconstructive valve operations.
of the leaflets is indicated in prolapsed leaflets secondary to localized ruptured chordae. Resection may also be used in localized fibrotic lesions. A quadrangular resection is the preferred method for the mural leaflet in order to avoid excess tension on the free edge of the leaflet. A triangular resection is mandatory on the aortic leaflet. The leaflet edges are sutured with interrupted 6-0 sutures . Leaflet resection is a simple, safe , and easy procedure which requires approximately 10 to 15 minutes . Chordal shortening. Chordal elongation has not been recognized as a major cause of mitral insufficiency until recently. 1 However, it does exist to a varying degree in 76% of the cases of so-called pure mitral insufficiency, whether of rheumatic or degenerative origin.' Chordal elongation is responsible for the prolapse of the leaflet into the atrium during systole. Chordal elongation of the anterior leaflet may be treated by either a shortening plastic repa ir of the chordae or a sliding pla stic repair of the papillary muscles. " It must be recognized that these techniques, especially chordal shortening, are difficult, time consuming, and may be hazardous unless extensive experience or adequate training is obtained on cadaver or animal hearts. Chordal fenestration and resection. Restricted motion of the leaflets is the result of commissural fusion, chordal fusion, chordal shortening , or chordal hypertrophy . Resection of secondary chordae attached to the ventricular surface of the mural leaflet results in improved mobility of leaflet tissue . Fenestration, i.e. , resection of a triangular portion of fused chordae, releases the subvalvular stenosis and improves leaflet motion . These techniques are not difficult but are
somewhat tedious and time consuming. It may be necessary to spend 15 to 20 minutes to mobilize the leaflet tissue adequately. These techniques most often are combined since associated lesions are common . For example , a complex condition such as floppy mitral valve associated with leaflet prolapse may require a quadrangular resection of the mural leaflet , chordal shortening, and a pro sthetic ring . Tricuspid repair. Tricuspid repair was achieved in all of the cases by remodelling of the anulus with a prosthetic ring." The prosthetic rings used until 1976 were continuous, rigid structures, whereas those used since 1976 have been open and flexible, designed to avoid injury to the bundle of His. Organic tricuspid disease necessitating commissurotomy was present in 51 cases (29%) . Other combined procedures are listed in Table IV . All patients undergoing mitral repair were subjected to an anticoagulation regimen for 2 months . Warfarin sodium (Coumadin), beginning 48 hours after operation, was the therapy used in those patients operated upon up to January, 1976. In the remaining patients , subcutaneous heparin was instituted the day following the operation and prolonged for 5 days , and then warfarin sodium was used for 2 months . Results The results have been analyzed by both percentage methods and actuarial methods with calculation of the confidence limits. 6 Deaths. There were 16 hospital deaths in 377 isolated mitral valve repairs (4.2%) and 25 hospital deaths
342
The Journal of Thoracic and Cardiovascular Surgery
Carpentier et al.
Table VD. Causes of reoperation
Table V. Causes of early deaths Ventricular failure Arrhythmia Myocardial infarction Hemorrhage Cerebral infarction Infection
17
Total
41
8 4 4 3 5
Residual mitral incompetence Recurrent mitral incompetence Mitral stenosis Hemolysis Thrombosis of atrium Endocarditis Aortic insufficiency Tricuspid insufficiency
Table VI. Causes of late deaths Cardiac failure Hemorrhage Cerebral hemorrhage Reoperation for residual or recurrent mitral incompetence Coronary artery disease Pulmonary hypertension Unknown Cancer Total
I .
3 I 3 9 2 I
4 I
24
in 174 mitral and tricuspid repairs (14%). The causes of early deaths are listed in Table V. Late deaths occurred in 19 of 265 patients followed in the mitral series (7%) and five of 76 patients followed in the mitral-tricuspid series (6.5%). The causes of late deaths are listed in Table VI. Fatal hemorrhage occurred in four patients receiving anticoagulants but in none of the patients not subjected to anticoagulation. Actuarial curves, hospital mortality rate included, show a survival rate of 82% at 9 years in the mitral series and a survival rate of 79% at 9 years in the mitral-tricuspid series (Fig. 3). If the hospital mortality rate is excluded, the survival rate is 88% at 9 years for the entire series. Correlation between survival and various factors, such as age, heart size, cause of disease, and pathophysiological type, showed a higher mortality rate in patients with a cardiothoracic ratio greater than 0.60 and a lower mortality rate in patients with type II incompetence. Thromboembolism. Among the 341 patients followed from 1 to 10 years (average 4.3 years), there were eight thromboembolic events in seven patients of the mitral series, for an embolic rate of 0.6% per patient-year, and five thromboembolic events in five patients in the mitral-tricuspid series, for an embolic rate of 0.8% per patient-year.. The actuarial curve shows a 92% incidence of thromboembolism-free patients at 7 years (Fig. 4). Seven of 12 thromboembolic events occurred in type III mitral incompetence and the
Total
First 2 years
9 3 3 2 0 I 0 0 18
Two to 10 years
3 9 4
o I
o I I 19
Table Vill. Technical factors related to reoperation Extensive resection of anterior leaflet Significant residual leak at operation Significant residual restricted motion of leaflet Error in ring size selection (stenosis) Detachment of prosthetic ring Extensive commissurotomy (hemolysis) Unrecognized prolapsed leaflet Poor indication Total
3 3 2 2 5 2 4 1
22/36
remaining in patients having had either a small prosthetic ring «30 mm) or a thrombosed or calcified left atrium. None of the thromboembolic episodes was fatal. All were transient. In one case, thrombosis of the left atrium necessitated reoperation and the outcome was successful. Forty-eight percent of the patients were without anticoagulation with the exception of the 2 month postoperative period and 41 % were in sinus rhythm. Reoperations. Reoperation was necessary in 26 patients in the mitral series (10%) and 11 patients in the mitral-tricuspid series (14%), for a total of 37 patients (11%). Almost half of the reoperations (18/37) were necessary within the first 2 years mainly because of residual mitral incompetence or stenosis (Table VII). This suggested that the causes were related to avoidable technical factors during operation. Technical faults could be clearly recognized in 22 cases of 37 at the time of reoperation. They are listed in Table VIII. No correlation was found between reoperations and such factors as pathophysiological type, lesions, age, and patient condition. Reoperations were relatively more frequent in the rheumatic valvular disease group than in the other groups. Fibroelastic deficiency associated with particularly thin chordae and type II mitral incompetence treated by shortening of the chordae did not pre-
Volume 79
MitraL incompetence
Number 3
March,1980
343
100".
90
80
.....--....----~
~-+--~--
70
I,
1m.
•
mitral
N = 377
'Y
mitral-tricuspid
N = 174
I
2
3
I
I
4
5
I 6
7
I
I
8
9
..
10Y_rs
Fig. 3. Actuarial survival curve. Hospital mortality rate is included. sent a higher incidence of reoperation than other groups or techniques. Recurrent tricuspid insufficiency or stenosis was never the cause of reoperation. There were five hospital deaths after reoperation for an incidence of 13%. Actuarial curves show that the probability of reoperation is 7% at 2 years and 13% at 8 years (Fig. 5). The rather flat curve between 2 and 8 years demonstrates the stability of an initially good result, with a rate of reoperation of I % per patient-year. Clinical status. Excluding late deaths and reoperations, 267 patients remained in whom the postoperative clinical status was assessed for more than I to 10 years (average 4.5 years). Preoperative and late postoperative N.Y.H.A. classifications are represented in Fig. 6. Only 10 patients (3.7%) returned to Class III after several years. Two patients (0.7%) returned to Class IV and will require reoperation within the next few months. The best results were in the group of patients with ruptured chordae and the poorest results in the group with mitral-tricuspid insufficiency. Auscultatory findings were normal in 147 patients, among whom five had a residual systolic murmur after operation which disappeared following reduction in heart size. One hundred thirteen patients had a mildto-moderate systolic murmur that remained unchanged from the time of operation. Seven patients had a moderate-to-severe systolic murmur, of which four
were recognized immediately after the operation. Two will require reoperation within the next year. Hemodynamic investigations. Hemodynamic studies were performed in 52 patients (Fig. 7). Twenty-one were performed in symptomatic patients having either a residual or a recurrent disability. In all cases, a close correlation was found between clinical and catheterization findings. Sixteen were performed in asymptomatic patients because of a persistent mild-to-moderate systolic murmur. Mean pulmonary pressures and capillary wedge pressures were within normal limits despite a mild-to-moderate amount of regurgitation documented by angiocardiography. Five were performed in asymptomatic patients with no residual murmur. The latter showed that reconstructive operation of the mitral valve can restore heart function to normal hemodynamic levels both at rest and during exercise. Discussion With the exception of ruptured chordae of the mural leaflet, in which reasonably good results were obtained by previous techniques.t" the various lesions responsible for mitral valve incompetence either were judged in the past to be unsuitable for repair or were treated by commissural plication'v " which failed to achieve predictable and stable long-term results. 14-18 The reason for this failure is that a single technique cannot
344
The Journal of Thoracic and Cardiovascular Surgery
Carpentier et al.
100%
------
90
80
70
I
2
I
3
,
I
5
4
I 6
I
I
7
8
I .. 9 Years
Fig. 4. Actuarial curve of thromboembolism-free patients.
Table IX. Indications and contraindications for mitral valve reconstruction Primary indication
Type I mitral incompetence Pure deformation of anulus dilatation Leaflet perforation < '4 .leaflet surface area > '4 leaflet surface area Acute endocarditis Type II mitral incompetence Ruptured chordae < '/3 mural leaflet < 'h mural leaflet > 'h mural leaflet Main chorda, anterior leaflet Paramedial chorda, anterior leaflet Elongated chordae < V3 mural leaflet > 'h mural leaflet Anterior leaflet Ruptured papillary muscle With contractile ventricular wall Noncontractile ventricular wall Type III mitral incompetence Valvular lesions Commissural fusion Leaflet thickening Associated subvalvular lesions Extensive Calcifications Localized Extensive
Relative indication
Contraindication
X X X X X X X X X X X X X X X X X X X X
Volume 79
Mitral incompetence
Number 3
March,1980
345
90
----,.. 80
70
T-+---+----+----+----+------l-+----+----+-__+_~ 2
I 3
4
I
I 5
6
I 7
I 8
I 9
I
..
10 Years
Fig. 5. Actuarial curve of reoperations.
correct multiple lesions. The breakthrough in reconstructive mitral valve surgery was the recognition that mitral insufficiency is almost never the result of one lesion but of multiple lesions which affect the various components of the mitral apparatus. I. 2 Therefore, reconstructive operations of the mitral valve should utilize several techniques directed toward the various lesions. I, 2, 4 Because of the complex mechanism of the mitral valve these techniques are of necessity complex, and this complexity may be viewed as a serious deterrent. This is compounded by the fact that currently available valvular bioprostheses offer advantages similar to reconstructive valve operations, i.e., low mortality rates, low thromboembolic rates and avoidance of anticoagulation, in addition to a simple and well-standardized technique. Another important factor limiting reconstructive valve operations is the psychological impact of a reoperation on the surgeon. Reoperation for a failing reconstructed valve causes the surgeon to feel some degree of guilt, whereas reoperation or even a fatal event complicating the course of a prosthetic valve is considered to be linked to the prosthesis itself. All of these factors may help to explain why surgeons react differently when faced with the same mitral valve lesion. In our institution, for example, the use of reconstructive mitral valve operation varies from 5% to 40% according to the individual surgeon. Most surgeons agree that indications exist for reconstruction,
207 II
13
51
III
110
10
IV
147
2
PRE-OP.
POST-OP.
Average follow up : 4 years 2 months Fig. 6. Preoperative and postoperative status of 270 patients (New York Heart Association classification).
but there are limitations to the techniques and both should be defined precisely. This brings us to the three problems raised at the beginning of this paper. Predictability of results. The predictability of the results obtained in this series and with this system of techniques may be regarded as superior to results obtained with the plication annuloplasty techniques, in which failures rates of 25% to 50% within a period of 4 to 8 years have been reported. 14-18 However, we con-
346
The Journal of Thoracic and Cardiovascular Surgery
Carpentier et al.
pre.
pre.
post.
~
30
30
25
25
20
20
15
15
25
25
20
">.
15
10
murmur ++ symptomatic N=21
pre.
post.
post.
20
yRCISE REST
murmur + asymptomatic N .. 16
15
10
~IS[ REST
no murmur asymptomatic N=5
Fig. 7. Hemodynamics in three series of patients. Left. Symptomatic patients with moderate-to-severe residual systolic murmur. Center, Asymptomatic patients with mild-to-moderate residual murmur. Right, Asymptomatic patients with no residual systolic murmur. sider that the valve failure rate of 13% at 8 years in this series remains too high, particularly when the expected failure rate of valvular bioprostheses at 10 years is 15% to 20%. If one takes into consideration that this series included all of the patients operated upon from the beginning of our experience, using techniques developed progressively, with uncertain case selection, one can expect better results in the future with more precise indications and a better knowledge of the limitations of the technique. Improvements in the predictability of the results can be achieved first by more accurate analysis of the mitral valve lesions both prior to and during operation. Four cases of residual mitral insufficiency in our series were due to unrecognized and therefore untreated elongation of the chordae. In one case, an excellent result was obtained by shortening the chordae at reoperation. Improvements may also be expected from more accurate testing of the repaired valve at operation. The technique proposed by Duran and Ubagc.!" of assessing the result on a beating heart with the left ventricle perfused by an arterial line, is interesting in this regard. If the result is not satisfactory, one should not hesitate to remove the valve. On reviewing the operative notes, we found that it was not rare in our series for the surgeon to leave a poor result because at the time he
did not know the prognosis of such an imperfect repair. By eliminating improper evaluation of the lesions, technical errors, and unsatisfactory repairs, one should obtain a failure rate of 1% per patient-year. Reproducibility of techniques. Reproducibility of techniques depends primarily on their simplicity. Though prosthetic ring implantation and leaflet resection-suture fit this requirement, the techniques of chordal fenestration and chordal shortening do not. The latter in particular may be regarded as too complex and therefore too hazardous for the expected benefit of the operation. For this reason, this technique has recently been simplified as follows (Fig. 8): A longitudinal incision involving the presenting half of the tip of the papillary muscle is performed to create a partial trench. One end of a double-armed suture is passed through one side of the trench. The other end is passed around the elongated chordae and then through the other side of the trench of the papillary muscle. Pulling the ends of the suture causes the excess length of the chordae to be buried in the papillary muscle. The suture is then tied, so that the trench is effectively closed around the buried portion of the chordae. An additional suture is placed through the papillary muscle to secure the closure. The degree of shortening is twice the distance
Volume 79
Mitral incompetence
Number 3 March. 1980
347
Fig. 8. Modified technique of chordal shortening. between the tip of the papillary muscle and the level at which the suture is passed through the papillary muscle. By modifying this distance one can modify the degree of shortening, which should correspond to the degree of prolapse of the leaflet. The time necessary to perform this procedure is approximately 10 minutes . Although the techniques have been simplified as much as possible, they continue to require an additional effort by the surgeon to acquire the necessary knowledge of the diseased mitral valve and to train himself on cadaver and animal hearts . At the time of operation, he should be prepared psychologically and technically to perform a valve repair rather than a valve replacement. Patient selection. Patient selection depends upon the type of valvular disease rather than the cause , the age, or the condition of the patient. Patient selection also depends upon the surgeon's experience and preference . Taking these two factors into consideration, the indication s for mitral valve reconstruction may be classified into three groups: primary indications, relative indications, and contraindications (Table IX). Primary indications. This group comprises lesions which can be treated easily and safely by reconstructive techniques, mainly anulus remodelling and leaflet resection-suture . The predictability of such repairs should approach 100%, unless one fails to recognize an
associated chordal elongation . A residual murmur of no hemodynamic significance may be present following anulus remodelling in type I mitral incompetence associated with a very large ventricle caused by excessive tension on the chordae. This murmur usually disappears after the heart has returned to a more normal size. Primary indications represent 10% to 20% of patients operated upon for mitral valve incompetence. Relative indications. This group comprises complicated lesions requiring complex techniques or multiple lesions requiring a combination of techniques, such as billowing mitral valve . These lesions can be repaired with predictable results after preliminary experience with the primary indications and adequate training on cadaver and animal hearts. Relative indications represent 20% to 30% of patients operated upon for mitral valve incompetence. Contraindi cations . With the exception of extensively calcified valves and markedly fibrotic valves, all mitral valve lesions were initially treated by conservative techniques to explore the possibilities and the limitations of this system of repair. As in all early experience , contraindications became progressively apparent, and they were not necessarily those we expected. For example, prolapse of anterior leaflets owing to ruptured chordae, which at first was thought to be a good indication, was found to be a contraindication when resection of one fourth of the leaflet or more was required . On the
The Journal of
348
Carpentier et al.
other hand, type III mitral incompetence proved to be associated with excellent and stable long-term results despite the severity of the lesions. Although affected by numerous early technical errors and inadequate indications, the results of this study confirm that this system of mitral valve repair compares favorably with other techniques of valve repair or replacement, giving a lower mortality rate, a lower incidence of thromboembolism, and better long-term functional results. The incidence of reoperation should be reduced further in the future by a better selection of patients. The additional effort by the surgeon to acquire the necessary expertise will be rewarded by a superior quality of life for the patients.
2 3
4
5
6
7
8
9
10
REFERENCES Carpentier A: Plastic and reconstructive mitral valve surgery, The Mitral Valve, D Kalmanson, ed., Acton, Mass., 1976, Publishing Sciences Group Carpentier A: Mitral valve reconstructive surgery, Operative Surgery, London, 1977, Butterworth & Co., Ltd. Carpentier A: La valvuloplastie reconstitutive. Une nouvelle technique de valvuloplastie mitrale. Presse Med 77:251, 1969 Carpentier A, Deloche A, Dauptain J, Soyer R, Blondeau P, Piwnica A, Dubost Ch: A new reconstructive operation for correction of mitral and tricuspid insufficiency. J THORAC CARDIOVASC SURG 61:1,1971 Carpentier A, Reiland J, Deloche A, Fabiani IN, D'Allaines CI, Blondeau Ph, Piwnica A, Chauvaud S, Dubost Ch: Conservative management of the prolapsed mitral valve. Ann Thorac Surg 26:294, 1978 Fabiani IN, Carpentier A: La methode actuarielle pour I,analyse statistique des resultats c1iniques et experimentaux. Nouv Presse Med 6:357, 1977 Ellis FH Jr, Frye RL, McGoon DC: Results of reconstructive operations for mitral insufficiency due to ruptured chordae tendineae. Surgery 59:165,1966 Gerbode F, Kerth WJ, Osborn 11, et al: Correction of mitral insufficiency by open operation. Ann Surg 155: 846, 1966 McGoon DC: Repair of mitral insufficiency due to ruptured chordae tendineae. J THORAC CARDtOVASC SURG 39:357, 1960 Lillehei CW, Gott VL, De Wall RA, et al: Surgical correction of mitral insufficiency by annuloplasty under direct vision. J Lancet 77:446, 1957
Thoracic and Cardiovascular Surgery
II Merendino KA, Thomas GI, Jesseph JE, et al: The open correction of rheumatic mitral regurgitation and/or stenosis. With special reference to regurgitation treated by posteromedial annuloplasty using a pump-oxygenator. Ann Surg 150:5, 1959 12 Reed GE, Tice DA, Clauss RH: Asymmetric exaggerated mitral annuloplasty. Repair of mitral insufficiency with hemodynamic predictability. J THORAC CARDlOV ASC SURG 49:752, 1965 13 Wooler GH, Nixon PGF, Grimshaw VA, et al: Experiences with the repair of the mitral valve in mitral incompetence. Thorax 17:49, 1962 14 Acar J, Caramanian M, Perrault M, et al: Les insuffisances mitrales par ruptures de cordages. Arch Mal Coeur 61:1724, 1968 15 Bigelow WG, Kuypers PJ, Heimbecker RO, et al: Clinical evaluation of the efficiency of mitral annuloplasty. Ann Surg 154:320, 1961 16 Bjork VO, Maiers E: Annuloplastic procedures for mitral insufficiency. Late results. J THORAC CARVIOVASC SURG 48:251, 1964 17 Dubost Ch: Evaluation of surgery for mitral valve disease. Am Heart J 82:143,1971 18 Penther PL Traitement chirurgical par annuloplastie des insuffisances mitrales. Resultats it long terme. Presse Med 77:1885, 1969 19 Duran CG, Ubago JLM: Conservative mitral valve surgery and developments in the technique of prosthetic ring annuloplasty, The Mitral Valve, D Kalmanson, ed., Acton, Mass., 1976, Publishing Sciences Group 20 Belcher JR: Evaluation of mitral annuloplasty for mitral regurgitaticn. Clinical and hemodynamic status four to forty-one months after surgery. Circulation 26:26, 1962 21 Cooley DA, Frazier OH, Norman JC: Mitral leaflet prolapse. Surgical treatment using a posterior annular collar prosthesis. Cardiovasc Dis Bull Texas Heart Inst 3:438, 1976 22 Oury JH, Forkerth TL, Hagan AD, et al: Indications and late results of reconstructive mitral surgery. Evaluation of the Carpentier ring, The Mitral Valve, D Kalmanson, ed., Acton, Mass., 1976, Publishing Sciences Group 23 Oury JH, Peterson KL, Forkerth TL, et al: Mitral valve replacement versus reconstruction. An analysis of indications and results of mitral valve procedures in a consecutive series of 80 patients. J THORAC CARDlOVASC SURG 73:825, 1977
(For discussion see page 354)
THORAC CARDIOVASC SURG
79:338-348, 1980
Reconstructive surgery of mitral valve incompetence Ten-year appraisal Between January, 1969, and January, 1978, 551 patients with mitral incompetence were treated by a system of reconstructive techniques. Mitral valve incompetence was classified into three types according to leaflet pliability; type 1 normal leaflet motion, 150 cases; type Il, leaflet prolapse, 213 cases; and type l1l, restricted leaflet motion, 188 cases. Associated tricuspid valvular disease was present in 174 cases (31.5%) and was treated by prosthetic ring annuloplasty. The operative mortality rate was 4.2% (16/377) in the mitral group and 14% (25/174) in the mitral-tricuspid group. Follow-up data are available in 341 patients from 1 year to 10 years (average 4 112 years). The late mortality rate was 7% (24/341). Actuarial curves including hospital mortality rate show an 82% survival rate at 9 years in the mitral group and a 79% rate in the mitral-tricuspid group. Thirty-seven patients (11%) underwent reoperation mainly for residual (17) or recurrent (16) mitral incompetence. Thromboembolism occurred in 12 patients for an embolic rate of 0.6% per patient-year, even though 48% were not given anticoagulants. A ccording to the New York Heart Association (N.Y.HA .) classification, 76% (207/270) of the patients were in Class I, 19% (51/270) were in Class ll, 4% (10/270) were in Class l1l, and 0.7% were in Class IV (2/270). Results of postoperative catheterization and angiocardiography are available in 52 patients. Comparison between the various groups shows that the best results were obtained in type II mitral incompetence, followed by type 1 and type III mitral incompetence. This experience demonstrates that predictable and stable long-term results have been achieved by techniques of valvular reconstruction with a low incidence of thromboembolism. Reproducibility of the techniques is a limiting factor which can be overcome by adequate training and progressive experience. Patient selection is based on the valvular disease rather than age, physical condition, or cause of valvular disease.
A. Carpentier, M.D. (by invitation), S. Chauvaud, M.D. (by invitation), J. N. Fabiani, M.D. (by invitation), A. Deloche, M.D. (by invitation),
J. ReIland, M.D. (by invitation), A. Lessana, M.D. (by invitation), Cl. d'Allaines, M.D. (by invitation), Ph. Blondeau, M.D. (by invitation), A. Piwnica, M.D. (by invitation), and Ch. Dubost, M.D., Paris, France
Reconstructive surgery of acquired mitral valve incompetence continues to raise three controversial problems: (l) predictability of results, (2) reproducibility of techniques, and (3) selection of patients. Having reviewed our experience with 551 mitral valve reconstructions performed at the Clinique de Chirurgie CarFrom the Chaire de Chirurgie Cardiovasculaire, Hopital Broussais, University of Paris, Paris, France. Supported by a grant from the C.N.R.S. Read at the Fifty-ninth Annual Meeting of the American Association for Thoracic Surgery, Boston, Mass., April 30 to May 2, 1979. Address for reprints: A. Carpentier, M.D., Hopital Broussais, 96 rue Didot, 75014-Paris (France).
338
diovasculaire of the Hopital Broussais in Paris, France, between January, 1969, and January, 1978, we will attempt to answer these points. Excluded from this study were 75 cases of congenital mitral valve disease and 286 cases of mitral valve incompetence associated with aortic valve replacement, which brings the total to 912 mitral valve reconstructions. Also excluded from the study were operations for pure mitral valve stenosis. Patients and methods During the period from January, 1969, to January, 1978, 551 patients with mitral valve incompetence were treated by reconstructive techniques. The ages of
0022-5223/80/030338+11$01.10/0 © 1980 The C. V. Mosby Co.
Volume 79
339
Mitral incompetence
Number 3
March,1980
Table I. Pathophysiological classification and valvular lesions* Normal leaflet motion Anulus dilatation Leaflet perforation Type II: Prolapsed leaflet Ruptured chordae Elongated chordae Ruptured papillary muscle Type III: Restricted leaflet motion Commissural fusion and leaflet thickening Associated fused chordae
Table III. Preoperative clinical condition of the patients
Type I:
159 157 2 213 109 103 I 179 52 127
'Calcifications were present in 26 cases.
Table II. Cause of valve lesions (in 419 cases) Cause
Rheumatic fever Endocarditis Myxoid degeneration (floppy valve) Fibroelastic deficiency Ischemic disease
~ 97 2 4 6
Type II
6I 10 49 30 I
IType III 159
the patients ranged from 2 to 77 years (average 37 years). Sixty patients were younger than 15 years of age. Pathophysiological classification. Mitral valve incompetence was classified into three types according to leaflet pliability!' 2; type I, normal leaflet motion; type II, prolapsed leaflet; and type III, restricted leaflet motion. Leaflet motion was assessed prior to operation by echocardiography or angiocardiography (or both) and during operation by mobilization of the leaflets with a nerve hook and by examination of the leaflets in the beating heart. In type I, the free edges of the leaflets remain below the plane of the anulus during systole and open normally during diastole. In type II, the free edge of one or both leaflets overrides the plane of the anulus during systole. In type III, one or both leaflets do not open fully during diastole. The classification is further defined as to the primary lesion. Table I categorizes the patients with regard to types and lesions. When associated lesions are present, the predominant lesion serves to classify the patient. For example, a patient with severe prolapse of the anterior leaflet owing to elongated chordae and some degree of restricted motion of the mural leaflet would be considered to have type II disease. The causes of the valve lesions were recognized in 419 patients and are listed in Table II. In nonrheumatic mitral valve insufficiency, we found it useful to introduce a clear distinction between fibroelastic deficiency and myxoid degeneration based on clinical patterns, gross structure, and histologic examination
N.Y.H.A. Class II N.Y.H.A. Class III N.Y.H.A. Class IV Acute rheumatic fever Angina Myocardial infarction Systemic embolism Giant atrium Sinus rhythm Atrial fibrillation Previous commissurotomy Previous plication of mitral anulus Moderate aortic insufficiency Tricuspid insufficiency
57 223 271 7
5 3 4 9 115 436 7
2 6 181
Legend: N.Y.H.A., New York Heart Association.
Table IV. Combined procedures used in 198 patients Atrial septal defect repair Coronary bypass graft Reduction of giant left atrium Resection of ventricular aneurysm Tricuspid ring annuloplasty
10 4 8 2
174
of the mitral valve (Fig. 1). Fibroelastic deficiency did not seem to be secondary to coronary artery disease, as the incidence of coronary artery disease in this series was small according to the age of the patients. Clinical condition of the patients. Patients were evaluated on the basis of the New York Heart Association (N.Y.H.A.) classification. The condition prior to operation is outlined in Table III. Patient selection for mitral valve repair was based on neither the clinical condition nor the age of the patient. Acute rheumatic fever associated with severe cardiac failure was not considered to be a contraindication, and six such patients were operated upon while receiving corticosteroid treatment. Indications were based upon the actual valvular lesions: All patients with noncalcified mitral valvular disease except those with acute bacterial endocarditis were considered to be candidates for valve repair. The percentage of noncalcified valves which were repaired varied from 15% to 80% according to the individual surgeon. Twenty-six patients with calcified mitral valvular disease were unexpectedly treated by reconstructive procedures because the lesions were found to be suitable for repair at operation. The duration of symptoms varied between 3 weeks and 18 years. Ten patients had previous commissurotomy and two had previous annuloplasties by the plication technique; combined procedures are listed in Table IV.
340 Carpentier et at.
The Journal of Thoracic and Cardiovascular Surgery
Fig. 1. Characteristic features in myxoid degeneration (Aj and fibroelastic deficiency (B j. Myxoid degeneration: Redundant, thickened, and yellowish leaflets with excess tissue. Severe dilatation of anulus. Elongated and irregular (thin or thickened) chordae. Disruption of collagen and elastic layers, and excess acid mucopolysaccharides. Average age at operation-48 years. Average time interval between first clinical symptom and operation-28 years. Fibroelastic deficiency (pellucid mitral valve): Smooth, thin, and translucent leaflets (no excess tissue). Moderate dilatation of anulus. Slightly elongated and very thin chordae. Fibroelastic and acid mucopolysaccharidedeficiency. Average age at operation-61 years. Average time interval between first clinical symptom and operation-6 years.
Surgical principles The following results are those obtained with a system of valvular reconstruction comprising four basic techniques that were used either individually or in association depending on the valvular lesions (Fig. 2): (l) anulus remodelling with a prosthetic ring, (2) leaflet tissue resection and repair, (3) chordal shortening, and (4) chordal fenestration and resection. Because technical details have been presented in previous papers, 1-5 we will recall here only the basic principles of these techniques and the occasional changes which have been introduced since previous publications. Anulus remodelling. Dilatation of the anulus fibrosus is recognized as one of the major factors in mitral insufficiency. The dilatation has certain characteristic features which must guide the technique of reconstruction: 1. Dilatation affects the mural leaflet and the commissural areas.
2. Dilatation does not affect the attachment of the aortic leaflet. 3. Dilatation is associated with a deformation of the anulus such that the anteroposterior diameter of the orifice is greater than the transverse diameter. 4. Dilatation is an evolving process which should be controlled by stabilizing the entire anulus. Prosthetic rings of a suitable shape and size are necessary to perform a measured annuloplasty which will restore the normal contour and thereby both a normal orificial area and optimum function of the valve. The selective flexibility of the prosthetic ring allows a more physiological repair. Prosthetic ring implantation does not present any surgical difficulty. It takes 20 to 30 minutes, depending on the surgeon's experience, and requires only that the spatial relationships between sutures on the anulus and on the prosthetic ring be carefully matched. Leaflet resection. Resection of abnormal segments
Volume 79
Mitral incompetence
Number 3
341
March . 1980
annulus remodelling
chordae shortening
leaflet
chordae
resect ion-suture
fenestration
Fig. 2. The four basic techniques used in reconstructive valve operations.
of the leaflets is indicated in prolapsed leaflets secondary to localized ruptured chordae. Resection may also be used in localized fibrotic lesions. A quadrangular resection is the preferred method for the mural leaflet in order to avoid excess tension on the free edge of the leaflet. A triangular resection is mandatory on the aortic leaflet. The leaflet edges are sutured with interrupted 6-0 sutures . Leaflet resection is a simple, safe , and easy procedure which requires approximately 10 to 15 minutes . Chordal shortening. Chordal elongation has not been recognized as a major cause of mitral insufficiency until recently. 1 However, it does exist to a varying degree in 76% of the cases of so-called pure mitral insufficiency, whether of rheumatic or degenerative origin.' Chordal elongation is responsible for the prolapse of the leaflet into the atrium during systole. Chordal elongation of the anterior leaflet may be treated by either a shortening plastic repa ir of the chordae or a sliding pla stic repair of the papillary muscles. " It must be recognized that these techniques, especially chordal shortening, are difficult, time consuming, and may be hazardous unless extensive experience or adequate training is obtained on cadaver or animal hearts. Chordal fenestration and resection. Restricted motion of the leaflets is the result of commissural fusion, chordal fusion, chordal shortening , or chordal hypertrophy . Resection of secondary chordae attached to the ventricular surface of the mural leaflet results in improved mobility of leaflet tissue . Fenestration, i.e. , resection of a triangular portion of fused chordae, releases the subvalvular stenosis and improves leaflet motion . These techniques are not difficult but are
somewhat tedious and time consuming. It may be necessary to spend 15 to 20 minutes to mobilize the leaflet tissue adequately. These techniques most often are combined since associated lesions are common . For example , a complex condition such as floppy mitral valve associated with leaflet prolapse may require a quadrangular resection of the mural leaflet , chordal shortening, and a pro sthetic ring . Tricuspid repair. Tricuspid repair was achieved in all of the cases by remodelling of the anulus with a prosthetic ring." The prosthetic rings used until 1976 were continuous, rigid structures, whereas those used since 1976 have been open and flexible, designed to avoid injury to the bundle of His. Organic tricuspid disease necessitating commissurotomy was present in 51 cases (29%) . Other combined procedures are listed in Table IV . All patients undergoing mitral repair were subjected to an anticoagulation regimen for 2 months . Warfarin sodium (Coumadin), beginning 48 hours after operation, was the therapy used in those patients operated upon up to January, 1976. In the remaining patients , subcutaneous heparin was instituted the day following the operation and prolonged for 5 days , and then warfarin sodium was used for 2 months . Results The results have been analyzed by both percentage methods and actuarial methods with calculation of the confidence limits. 6 Deaths. There were 16 hospital deaths in 377 isolated mitral valve repairs (4.2%) and 25 hospital deaths
342
The Journal of Thoracic and Cardiovascular Surgery
Carpentier et al.
Table VD. Causes of reoperation
Table V. Causes of early deaths Ventricular failure Arrhythmia Myocardial infarction Hemorrhage Cerebral infarction Infection
17
Total
41
8 4 4 3 5
Residual mitral incompetence Recurrent mitral incompetence Mitral stenosis Hemolysis Thrombosis of atrium Endocarditis Aortic insufficiency Tricuspid insufficiency
Table VI. Causes of late deaths Cardiac failure Hemorrhage Cerebral hemorrhage Reoperation for residual or recurrent mitral incompetence Coronary artery disease Pulmonary hypertension Unknown Cancer Total
I .
3 I 3 9 2 I
4 I
24
in 174 mitral and tricuspid repairs (14%). The causes of early deaths are listed in Table V. Late deaths occurred in 19 of 265 patients followed in the mitral series (7%) and five of 76 patients followed in the mitral-tricuspid series (6.5%). The causes of late deaths are listed in Table VI. Fatal hemorrhage occurred in four patients receiving anticoagulants but in none of the patients not subjected to anticoagulation. Actuarial curves, hospital mortality rate included, show a survival rate of 82% at 9 years in the mitral series and a survival rate of 79% at 9 years in the mitral-tricuspid series (Fig. 3). If the hospital mortality rate is excluded, the survival rate is 88% at 9 years for the entire series. Correlation between survival and various factors, such as age, heart size, cause of disease, and pathophysiological type, showed a higher mortality rate in patients with a cardiothoracic ratio greater than 0.60 and a lower mortality rate in patients with type II incompetence. Thromboembolism. Among the 341 patients followed from 1 to 10 years (average 4.3 years), there were eight thromboembolic events in seven patients of the mitral series, for an embolic rate of 0.6% per patient-year, and five thromboembolic events in five patients in the mitral-tricuspid series, for an embolic rate of 0.8% per patient-year.. The actuarial curve shows a 92% incidence of thromboembolism-free patients at 7 years (Fig. 4). Seven of 12 thromboembolic events occurred in type III mitral incompetence and the
Total
First 2 years
9 3 3 2 0 I 0 0 18
Two to 10 years
3 9 4
o I
o I I 19
Table Vill. Technical factors related to reoperation Extensive resection of anterior leaflet Significant residual leak at operation Significant residual restricted motion of leaflet Error in ring size selection (stenosis) Detachment of prosthetic ring Extensive commissurotomy (hemolysis) Unrecognized prolapsed leaflet Poor indication Total
3 3 2 2 5 2 4 1
22/36
remaining in patients having had either a small prosthetic ring «30 mm) or a thrombosed or calcified left atrium. None of the thromboembolic episodes was fatal. All were transient. In one case, thrombosis of the left atrium necessitated reoperation and the outcome was successful. Forty-eight percent of the patients were without anticoagulation with the exception of the 2 month postoperative period and 41 % were in sinus rhythm. Reoperations. Reoperation was necessary in 26 patients in the mitral series (10%) and 11 patients in the mitral-tricuspid series (14%), for a total of 37 patients (11%). Almost half of the reoperations (18/37) were necessary within the first 2 years mainly because of residual mitral incompetence or stenosis (Table VII). This suggested that the causes were related to avoidable technical factors during operation. Technical faults could be clearly recognized in 22 cases of 37 at the time of reoperation. They are listed in Table VIII. No correlation was found between reoperations and such factors as pathophysiological type, lesions, age, and patient condition. Reoperations were relatively more frequent in the rheumatic valvular disease group than in the other groups. Fibroelastic deficiency associated with particularly thin chordae and type II mitral incompetence treated by shortening of the chordae did not pre-
Volume 79
MitraL incompetence
Number 3
March,1980
343
100".
90
80
.....--....----~
~-+--~--
70
I,
1m.
•
mitral
N = 377
'Y
mitral-tricuspid
N = 174
I
2
3
I
I
4
5
I 6
7
I
I
8
9
..
10Y_rs
Fig. 3. Actuarial survival curve. Hospital mortality rate is included. sent a higher incidence of reoperation than other groups or techniques. Recurrent tricuspid insufficiency or stenosis was never the cause of reoperation. There were five hospital deaths after reoperation for an incidence of 13%. Actuarial curves show that the probability of reoperation is 7% at 2 years and 13% at 8 years (Fig. 5). The rather flat curve between 2 and 8 years demonstrates the stability of an initially good result, with a rate of reoperation of I % per patient-year. Clinical status. Excluding late deaths and reoperations, 267 patients remained in whom the postoperative clinical status was assessed for more than I to 10 years (average 4.5 years). Preoperative and late postoperative N.Y.H.A. classifications are represented in Fig. 6. Only 10 patients (3.7%) returned to Class III after several years. Two patients (0.7%) returned to Class IV and will require reoperation within the next few months. The best results were in the group of patients with ruptured chordae and the poorest results in the group with mitral-tricuspid insufficiency. Auscultatory findings were normal in 147 patients, among whom five had a residual systolic murmur after operation which disappeared following reduction in heart size. One hundred thirteen patients had a mildto-moderate systolic murmur that remained unchanged from the time of operation. Seven patients had a moderate-to-severe systolic murmur, of which four
were recognized immediately after the operation. Two will require reoperation within the next year. Hemodynamic investigations. Hemodynamic studies were performed in 52 patients (Fig. 7). Twenty-one were performed in symptomatic patients having either a residual or a recurrent disability. In all cases, a close correlation was found between clinical and catheterization findings. Sixteen were performed in asymptomatic patients because of a persistent mild-to-moderate systolic murmur. Mean pulmonary pressures and capillary wedge pressures were within normal limits despite a mild-to-moderate amount of regurgitation documented by angiocardiography. Five were performed in asymptomatic patients with no residual murmur. The latter showed that reconstructive operation of the mitral valve can restore heart function to normal hemodynamic levels both at rest and during exercise. Discussion With the exception of ruptured chordae of the mural leaflet, in which reasonably good results were obtained by previous techniques.t" the various lesions responsible for mitral valve incompetence either were judged in the past to be unsuitable for repair or were treated by commissural plication'v " which failed to achieve predictable and stable long-term results. 14-18 The reason for this failure is that a single technique cannot
344
The Journal of Thoracic and Cardiovascular Surgery
Carpentier et al.
100%
------
90
80
70
I
2
I
3
,
I
5
4
I 6
I
I
7
8
I .. 9 Years
Fig. 4. Actuarial curve of thromboembolism-free patients.
Table IX. Indications and contraindications for mitral valve reconstruction Primary indication
Type I mitral incompetence Pure deformation of anulus dilatation Leaflet perforation < '4 .leaflet surface area > '4 leaflet surface area Acute endocarditis Type II mitral incompetence Ruptured chordae < '/3 mural leaflet < 'h mural leaflet > 'h mural leaflet Main chorda, anterior leaflet Paramedial chorda, anterior leaflet Elongated chordae < V3 mural leaflet > 'h mural leaflet Anterior leaflet Ruptured papillary muscle With contractile ventricular wall Noncontractile ventricular wall Type III mitral incompetence Valvular lesions Commissural fusion Leaflet thickening Associated subvalvular lesions Extensive Calcifications Localized Extensive
Relative indication
Contraindication
X X X X X X X X X X X X X X X X X X X X
Volume 79
Mitral incompetence
Number 3
March,1980
345
90
----,.. 80
70
T-+---+----+----+----+------l-+----+----+-__+_~ 2
I 3
4
I
I 5
6
I 7
I 8
I 9
I
..
10 Years
Fig. 5. Actuarial curve of reoperations.
correct multiple lesions. The breakthrough in reconstructive mitral valve surgery was the recognition that mitral insufficiency is almost never the result of one lesion but of multiple lesions which affect the various components of the mitral apparatus. I. 2 Therefore, reconstructive operations of the mitral valve should utilize several techniques directed toward the various lesions. I, 2, 4 Because of the complex mechanism of the mitral valve these techniques are of necessity complex, and this complexity may be viewed as a serious deterrent. This is compounded by the fact that currently available valvular bioprostheses offer advantages similar to reconstructive valve operations, i.e., low mortality rates, low thromboembolic rates and avoidance of anticoagulation, in addition to a simple and well-standardized technique. Another important factor limiting reconstructive valve operations is the psychological impact of a reoperation on the surgeon. Reoperation for a failing reconstructed valve causes the surgeon to feel some degree of guilt, whereas reoperation or even a fatal event complicating the course of a prosthetic valve is considered to be linked to the prosthesis itself. All of these factors may help to explain why surgeons react differently when faced with the same mitral valve lesion. In our institution, for example, the use of reconstructive mitral valve operation varies from 5% to 40% according to the individual surgeon. Most surgeons agree that indications exist for reconstruction,
207 II
13
51
III
110
10
IV
147
2
PRE-OP.
POST-OP.
Average follow up : 4 years 2 months Fig. 6. Preoperative and postoperative status of 270 patients (New York Heart Association classification).
but there are limitations to the techniques and both should be defined precisely. This brings us to the three problems raised at the beginning of this paper. Predictability of results. The predictability of the results obtained in this series and with this system of techniques may be regarded as superior to results obtained with the plication annuloplasty techniques, in which failures rates of 25% to 50% within a period of 4 to 8 years have been reported. 14-18 However, we con-
346
The Journal of Thoracic and Cardiovascular Surgery
Carpentier et al.
pre.
pre.
post.
~
30
30
25
25
20
20
15
15
25
25
20
">.
15
10
murmur ++ symptomatic N=21
pre.
post.
post.
20
yRCISE REST
murmur + asymptomatic N .. 16
15
10
~IS[ REST
no murmur asymptomatic N=5
Fig. 7. Hemodynamics in three series of patients. Left. Symptomatic patients with moderate-to-severe residual systolic murmur. Center, Asymptomatic patients with mild-to-moderate residual murmur. Right, Asymptomatic patients with no residual systolic murmur. sider that the valve failure rate of 13% at 8 years in this series remains too high, particularly when the expected failure rate of valvular bioprostheses at 10 years is 15% to 20%. If one takes into consideration that this series included all of the patients operated upon from the beginning of our experience, using techniques developed progressively, with uncertain case selection, one can expect better results in the future with more precise indications and a better knowledge of the limitations of the technique. Improvements in the predictability of the results can be achieved first by more accurate analysis of the mitral valve lesions both prior to and during operation. Four cases of residual mitral insufficiency in our series were due to unrecognized and therefore untreated elongation of the chordae. In one case, an excellent result was obtained by shortening the chordae at reoperation. Improvements may also be expected from more accurate testing of the repaired valve at operation. The technique proposed by Duran and Ubagc.!" of assessing the result on a beating heart with the left ventricle perfused by an arterial line, is interesting in this regard. If the result is not satisfactory, one should not hesitate to remove the valve. On reviewing the operative notes, we found that it was not rare in our series for the surgeon to leave a poor result because at the time he
did not know the prognosis of such an imperfect repair. By eliminating improper evaluation of the lesions, technical errors, and unsatisfactory repairs, one should obtain a failure rate of 1% per patient-year. Reproducibility of techniques. Reproducibility of techniques depends primarily on their simplicity. Though prosthetic ring implantation and leaflet resection-suture fit this requirement, the techniques of chordal fenestration and chordal shortening do not. The latter in particular may be regarded as too complex and therefore too hazardous for the expected benefit of the operation. For this reason, this technique has recently been simplified as follows (Fig. 8): A longitudinal incision involving the presenting half of the tip of the papillary muscle is performed to create a partial trench. One end of a double-armed suture is passed through one side of the trench. The other end is passed around the elongated chordae and then through the other side of the trench of the papillary muscle. Pulling the ends of the suture causes the excess length of the chordae to be buried in the papillary muscle. The suture is then tied, so that the trench is effectively closed around the buried portion of the chordae. An additional suture is placed through the papillary muscle to secure the closure. The degree of shortening is twice the distance
Volume 79
Mitral incompetence
Number 3 March. 1980
347
Fig. 8. Modified technique of chordal shortening. between the tip of the papillary muscle and the level at which the suture is passed through the papillary muscle. By modifying this distance one can modify the degree of shortening, which should correspond to the degree of prolapse of the leaflet. The time necessary to perform this procedure is approximately 10 minutes . Although the techniques have been simplified as much as possible, they continue to require an additional effort by the surgeon to acquire the necessary knowledge of the diseased mitral valve and to train himself on cadaver and animal hearts . At the time of operation, he should be prepared psychologically and technically to perform a valve repair rather than a valve replacement. Patient selection. Patient selection depends upon the type of valvular disease rather than the cause , the age, or the condition of the patient. Patient selection also depends upon the surgeon's experience and preference . Taking these two factors into consideration, the indication s for mitral valve reconstruction may be classified into three groups: primary indications, relative indications, and contraindications (Table IX). Primary indications. This group comprises lesions which can be treated easily and safely by reconstructive techniques, mainly anulus remodelling and leaflet resection-suture . The predictability of such repairs should approach 100%, unless one fails to recognize an
associated chordal elongation . A residual murmur of no hemodynamic significance may be present following anulus remodelling in type I mitral incompetence associated with a very large ventricle caused by excessive tension on the chordae. This murmur usually disappears after the heart has returned to a more normal size. Primary indications represent 10% to 20% of patients operated upon for mitral valve incompetence. Relative indications. This group comprises complicated lesions requiring complex techniques or multiple lesions requiring a combination of techniques, such as billowing mitral valve . These lesions can be repaired with predictable results after preliminary experience with the primary indications and adequate training on cadaver and animal hearts. Relative indications represent 20% to 30% of patients operated upon for mitral valve incompetence. Contraindi cations . With the exception of extensively calcified valves and markedly fibrotic valves, all mitral valve lesions were initially treated by conservative techniques to explore the possibilities and the limitations of this system of repair. As in all early experience , contraindications became progressively apparent, and they were not necessarily those we expected. For example, prolapse of anterior leaflets owing to ruptured chordae, which at first was thought to be a good indication, was found to be a contraindication when resection of one fourth of the leaflet or more was required . On the
The Journal of
348
Carpentier et al.
other hand, type III mitral incompetence proved to be associated with excellent and stable long-term results despite the severity of the lesions. Although affected by numerous early technical errors and inadequate indications, the results of this study confirm that this system of mitral valve repair compares favorably with other techniques of valve repair or replacement, giving a lower mortality rate, a lower incidence of thromboembolism, and better long-term functional results. The incidence of reoperation should be reduced further in the future by a better selection of patients. The additional effort by the surgeon to acquire the necessary expertise will be rewarded by a superior quality of life for the patients.
2 3
4
5
6
7
8
9
10
REFERENCES Carpentier A: Plastic and reconstructive mitral valve surgery, The Mitral Valve, D Kalmanson, ed., Acton, Mass., 1976, Publishing Sciences Group Carpentier A: Mitral valve reconstructive surgery, Operative Surgery, London, 1977, Butterworth & Co., Ltd. Carpentier A: La valvuloplastie reconstitutive. Une nouvelle technique de valvuloplastie mitrale. Presse Med 77:251, 1969 Carpentier A, Deloche A, Dauptain J, Soyer R, Blondeau P, Piwnica A, Dubost Ch: A new reconstructive operation for correction of mitral and tricuspid insufficiency. J THORAC CARDIOVASC SURG 61:1,1971 Carpentier A, Reiland J, Deloche A, Fabiani IN, D'Allaines CI, Blondeau Ph, Piwnica A, Chauvaud S, Dubost Ch: Conservative management of the prolapsed mitral valve. Ann Thorac Surg 26:294, 1978 Fabiani IN, Carpentier A: La methode actuarielle pour I,analyse statistique des resultats c1iniques et experimentaux. Nouv Presse Med 6:357, 1977 Ellis FH Jr, Frye RL, McGoon DC: Results of reconstructive operations for mitral insufficiency due to ruptured chordae tendineae. Surgery 59:165,1966 Gerbode F, Kerth WJ, Osborn 11, et al: Correction of mitral insufficiency by open operation. Ann Surg 155: 846, 1966 McGoon DC: Repair of mitral insufficiency due to ruptured chordae tendineae. J THORAC CARDtOVASC SURG 39:357, 1960 Lillehei CW, Gott VL, De Wall RA, et al: Surgical correction of mitral insufficiency by annuloplasty under direct vision. J Lancet 77:446, 1957
Thoracic and Cardiovascular Surgery
II Merendino KA, Thomas GI, Jesseph JE, et al: The open correction of rheumatic mitral regurgitation and/or stenosis. With special reference to regurgitation treated by posteromedial annuloplasty using a pump-oxygenator. Ann Surg 150:5, 1959 12 Reed GE, Tice DA, Clauss RH: Asymmetric exaggerated mitral annuloplasty. Repair of mitral insufficiency with hemodynamic predictability. J THORAC CARDlOV ASC SURG 49:752, 1965 13 Wooler GH, Nixon PGF, Grimshaw VA, et al: Experiences with the repair of the mitral valve in mitral incompetence. Thorax 17:49, 1962 14 Acar J, Caramanian M, Perrault M, et al: Les insuffisances mitrales par ruptures de cordages. Arch Mal Coeur 61:1724, 1968 15 Bigelow WG, Kuypers PJ, Heimbecker RO, et al: Clinical evaluation of the efficiency of mitral annuloplasty. Ann Surg 154:320, 1961 16 Bjork VO, Maiers E: Annuloplastic procedures for mitral insufficiency. Late results. J THORAC CARVIOVASC SURG 48:251, 1964 17 Dubost Ch: Evaluation of surgery for mitral valve disease. Am Heart J 82:143,1971 18 Penther PL Traitement chirurgical par annuloplastie des insuffisances mitrales. Resultats it long terme. Presse Med 77:1885, 1969 19 Duran CG, Ubago JLM: Conservative mitral valve surgery and developments in the technique of prosthetic ring annuloplasty, The Mitral Valve, D Kalmanson, ed., Acton, Mass., 1976, Publishing Sciences Group 20 Belcher JR: Evaluation of mitral annuloplasty for mitral regurgitaticn. Clinical and hemodynamic status four to forty-one months after surgery. Circulation 26:26, 1962 21 Cooley DA, Frazier OH, Norman JC: Mitral leaflet prolapse. Surgical treatment using a posterior annular collar prosthesis. Cardiovasc Dis Bull Texas Heart Inst 3:438, 1976 22 Oury JH, Forkerth TL, Hagan AD, et al: Indications and late results of reconstructive mitral surgery. Evaluation of the Carpentier ring, The Mitral Valve, D Kalmanson, ed., Acton, Mass., 1976, Publishing Sciences Group 23 Oury JH, Peterson KL, Forkerth TL, et al: Mitral valve replacement versus reconstruction. An analysis of indications and results of mitral valve procedures in a consecutive series of 80 patients. J THORAC CARDlOVASC SURG 73:825, 1977
(For discussion see page 354)