Mitral regurgitation due to ruptured chordae tendineae

Mitral regurgitation due to ruptured chordae tendineae

J THoRAc CARDIOVASC SURG 89:491-498, 1985 Mitral regurgitation due to ruptured chordae tendineae Early and late results of valve repair From 1958 t...

892KB Sizes 0 Downloads 5 Views

J

THoRAc CARDIOVASC SURG

89:491-498, 1985

Mitral regurgitation due to ruptured chordae tendineae Early and late results of valve repair From 1958 through 1980, 131 patients had repair of ruptured chordae tendineae of the mitral valve; 62 % were men. Ages ranged from 5 to 70 years (median 57). Chordae to the anterior mitral leaflet were ruptured in 44 patients (34%), to the posterior mitral leaflet in 85 (65%), and to both leaflets in two patients (1%). The mitral valve was repaired by leaflet plication without resection in 116 patients, plication after wedge resection of the unsupported leaflet in six, Ivalon sponge buttress of the posterior leaflet in three, resuspension of chordae in two, and annuloplasty alone in the remaining four. Mitral valve annuloplasty was performed in addition to leaflet repair in 115 patients (88 %). Operative «30 days) mortality was 6.1 %. Survival rate of patients dismissed from the hospital was 92 % at 5 years and 73 % at 10 years. There were no differences in late survival or risk of reoperation for recurrent or residual mitral insufficiency between patients with ruptured chordae to the anterior leaflet and those with ruptured chordae to the posterior leaflet. Survival was significantly better for the group with repair than it was for a group that underwent mitral valvereplacement for ruptured chordae during this same time interval (5 year survival rate, 92 % versus 72 %, p < 0.003). The incidence of thromboembolism after repair was 1.8 episodes/100 patient-years compared with 8.0 episodes/100 patient-years after replacement. Our data indicate that valvuloplasty is the procedure of choice for most patients with mitral regurgitation owing to ruptured chordae tendineae, including selected patients with ruptured chordae to the anterior: leaflet.

Thomas A. Orszulak, M.D. (by invitation), Hartzell V. Schaff, M.D. (by invitation), Gordon K. Danielson, M.D., Jeffrey M. Piehler, M.D. (by invitation), James R. Pluth, M.D., Robert L. Frye, M.D. (by invitation), and Dwight C. McGoon, M.D., Rochester, Minn. Statistical Consultant: Lila R. Elveback, Ph.D.

Rupture of chordae tendineae is the cause of valve incompetence in approximately 1 of 5 patients with clinically important mitral insufficiency.I Chordal rupture may result from bacterial endocarditis, rheumatic valvulitis, and occasionally chest traumaY Some patients have associated myxomatous degeneration of the valve, whereas others have no obvious associated cardiac disease.' Operative treatment of ruptured chordaeis indicated in patients who have serious functional From the Section of Thoracic and Cardiovascular Surgery and the Division of Cardiovascular Diseases and Internal Medicine, Mayo Clinic and Mayo Foundation, Rochester, Minn.. Read at the Sixty-fourth Annual Meeting of The American Association for Thoracic Surgery, New York, N. Y. May 7-9, 1984. Address for reprints: T. A. Orszulak, M.D., Mayo Clinic, 200 First Street SW, Rochester, Minn. 55905.

limitation from congestive heart failure and in asymptomatic patients who have progressive left ventricular dysfunction documented by noninvasive or invasive tests. Historically, repair with or without prosthetic materials was the first method used to restore mitral valve competence in patients with insufficiency resulting from ruptured chordae.f Although the operative risk for mitral valve repair is low and the risk of late thromboembolism is minimal,"!' some surgeons have advocated prosthetic replacement of the regurgitant mitral valve with ruptured chordae tendineae.v'<" Proponents of valve replacement question the durability of valve reconstruction and emphasize that competence can be restored with more certainty with a prosthesis. To address these and other issues, we reviewed our entire experience with cardiac operations for mitral valve 491

The Journal of Thoracic and Cardiovascular Surgery

4 9 2 Orszufak et af.

Table I. Patients with mitral regurgitation due to rupture of chordae tendineae (Mayo Clinic 1958 to 1980) Mitral valve repair No. of patients Median age (yr) No. with angina Early «30 days) mortality Incidence of late thromboembolism (episodea/ l Ot) patient-years) Five-year survival (%)

131 57 36 (27%) 8 (6.1%) 1.8

88t

Mitral valve replacement 106'

64 19 (18%) 8 (7.5%) 8.0 72

*In seven patients, valve repair was unsatisfactory and prosthetic valve replacement was necessary at the same operation. tSignificantly different from group with mitral valve replacement (p

< 0.003).

insufficiency primarily resulting from ruptured chordae tendineae and focused on the early and late clinical results after valve repair. Our series From 1958 through 1980, 237 patients underwent correction of mitral regurgitation owing to ruptured chordae tendineae. Of these, 131 had mitral valve repair and 106 had mitral valve replacement (Table I). The late results of the 82 men and 49 women who had mitral valve repair form the basisof this report; those undergoing mitral valve replacement are analyzed for perspective. The ages of the group with repair ranged from 5 to 70 years (median 57). All patients undergoing repair were symptomatic at operation, and 79% had New York Heart Association Class III or IV disability. Thirty-six patients (27%) had angina pectoris, and 23 (18%) had a history of bacterial endocarditis. The duration of preoperative symptoms and murmurs was variable. Fortyseven patients had significant congestive heart failure from mitral regurgitation for less than 9 months preoperatively; 29 of these patients had a longer history of heart murmur. Sixty-eight patients (52%) had auscultatory findings of mitral regurgitation for more than 8 years preoperatively. More than two thirds of the patients were referred for mitral valve operation without preoperative cardiac catheterization." In these patients, the preoperative diagnosis was established by clinical findings, and operation was recommended because of progressive functional limitation resulting from congestive heart failure. Methods Clinical records and operative reports were reviewed with particular attention to the mitral valve disease and

the method of repair. All deaths within 30 days of operation were considered as operative mortality. Subsequent follow-up by routine clinic visit or direct correspondence with the patient or his physician (or both) was complete for 98.5% of the patients. The influence of multiple preoperative and intraoperative factors on late survival was assessed by univariate analysis. Multivariate analysis was used to test the independence of risk factors. Late survival was expressed by the KaplanMeier method and was stratified according to multiple patient variables. The individual survival curves were compared by the log rank test. Operative findings Surgical technique varied during the 23 year period under review. Cardiopulmonary bypass with mild (30 0 C to 32 0 C) systemic hypothermia was used for most patients, and exposure of the mitral valve was facilitated with intermittent aortic cross-clamping. Since 1978, we have used cardioplegic arrest combined with topical hypothermia for myocardial protection. Mitral valve repair was accomplished by leaflet plication technique? without excision of valve tissue in 116 patients and by wedge excision of the unsupported segment with leaflet repair in six patients.'? Three patients with ruptured chordae to the posterior mitral valve leaflet had valve repair by buttress of the leaflet with an Ivalon sponge.I? One patient with ruptured chordae to both leaflets had fixation of the posterior leaflet with Ivalon sponge and plication of the anterior leaflet, and a second patient with ruptured chordae to both leaflets had Ivalon sponge buttress of the posterior leaflet and had repair of the anterior leaflet effected by triangular leaflet excision and reapproximation. In the remaining four patients, the ruptured chordae were adjacent to the posteromedial commissure, and mitral valve competence was restored by annuloplasty alone. Patients were categorized according to the location of the ruptured chordae. Rupture of chordae to the anterior (aortic) leaflet was identified in 44 patients (34%) and rupture of chordae to the posterior (mural) leaflet in 85 patients (65%). Two patients (1%) had chordae to both anterior and posterior leaflets ruptured. Additional mitral valve abnormalities were observed in 55 patients. Significant prolapse of the anterior or posterior leaflet or both occurred in 29 patients, and a natural (congenital) valve cleft was identified in 23 additional patients. Three patients had papillary muscle rupture in addition to ruptured chordae tendineae. In addition to having direct repair of the leaflet in the area of the ruptured chordae, 23 patients had suture closure ofvalve clefts. Valve repair was supplemented by annuloplasty using plication of both posteromedial and

Volume 89 Number 4

Mitral regurgitation 4 9 3

April, 1985

Table II. Early deaths «30 days) after mitral valve repair for ruptured chordae* Sex and age (yr)

Date of operation

Cardiac rhythm

Location of ruptured chordae

2

F, 49 F, 37

8-1-58 8-21-59

AF AF

Posterior leaflet Anterior leaflet

3

M,56

11-7-60

AF

Anterior leaflet

4

M,21

2-7-61

Sinus

Anterior leaflet

5

F, 56

4-24-61

AF

6

M,61

7-31-61

7

M,60

8

F, 30

Case I

*AF,

Associated mitral valve disease

Associated cardiac procedure

None Anterior leaflet vegetation None

None Debridement of vegetation Aortic valve repair

Posterior leaflet

Aortic insufficiency; perforation of anterior leaflet None

Pericardial patch closure of leaflet perforation None

AF

Posterior leaflet

None

None

11-11-64

AF

Posterior leaflet

None

None

6-7-67

AF

Anterior leaflet

None

None

Cause of death (interval from operation) Stroke (7 days) Persistent CHF (13 days) Low cardiac output (intraoperatively) Low cardiac output (3 hr) Stroke, air embolism (12 hr) Persistent CHF (16 days) Sudden arrest (3 days) Stroke (7 days)

Atrial fibrillation. CH F, Congestive heart failure.

anterolateral commissures in 37 patients, plication of the posteromedial commissure alone in 69, and plication of the anterolateral commissure alone in four and by insertion of a prosthetic ring (Carpentier) in five. Early in our experience, the intraoperative assessment of adequacy of repair was limited to visual analysis through the opened atrium after the injection of saline into the left ventricle. This was supplemented by observing for left atrial 'V' waves and palpable thrills after bypass. Our present evaluation includes the preliminary injection of saline, but the quantitative assessment is improved with the use of indocyanine green doublesampling dye curves obtained after discontinuing heartlung bypass. Associated procedures were performed in 22 patients at operative repair. Six patients underwent concomitant aortic valve repair or replacement, six had closure of an atrial septal defect, four had tricuspid valve annuloplasty, three underwent simultaneous coronary artery revascularization, and one each had left ventricular apical aneurysm plication, ascending aortic aneurysmorrhaphy, or permanent pacemaker insertion.

Early and late results Eight patients died within 30 days of operation, for an operative mortality of 6.1%. Incomplete correction of mitral regurgitation probably contributed to the deaths of Patients 2, 3, 4 and 6 (Table II). Stroke, presumably embolic, was the cause of death in three patients. One patient remained comatose after mitral valve repair and died 12 hours postoperatively. The other two patients had sudden cerebrovascular accidents 7 days postoperatively. The eighth operative death was that of a 60-

year-old man who had repair of ruptured chordae to the posterior leaflet. He died suddenly on the third postoperative day, presumably of an arrhythmia. Of the eight patients who died early after mitral valve repair for ruptured chordae, seven were in atrial fibrillation preoperatively, including the two patients who had stroke 7 days postoperatively. Five of the eight deaths occurred before a reliable mechanical heart valve prosthesis was available. (At our institution the first StarrEdwards heart valve was implanted on April 29, 1961, by J. W. Kirklin, M.D.) Since 1965, 96 patients have undergone repair of ruptured chordae (28 patients had ruptured chordae to the anterior leaflet), and there has been only one early death, for an operative mortality of 1%. A number of nonfatal complications occurred after repair: 28 patients had new postoperative atrial arrhythmias (primarily atrial fibrillation and atrial flutter), seven had respiratory insufficiency necessitating tracheostomy and intensive pulmonary physiotherapy, five had persistent mediastinal hemorrhage that necessitated reoperation, four had nonfatal cerebrovascular accidents, three had atrioventricular block, two had myocardial infarction, two had sternal infection, and one had renal failure. At dismissal after repair, 76 patients (58%) had murmurs of mitral regurgitation: 31 (41%) had anterior leaflet repair and 45 (59%) had posterior leaflet repair. Fifty-seven patients died during the follow-up period. In 27 patients, death was considered cardiac in origin, and the causes included congestive heart failure, myocardial infarction, and subsequent reoperation. The late deaths of 20 patients were noncardiac in origin. The

The Journal of Thoracic and Cardiovascular Surgery

4 9 4 Orszulak et al.

;§!.

100 90(121)

~

80

>

70

...

;;l '0

(105)

Cll

>

60 40

:0

30

Cll .0

20

a::

10

... Repair ruptured chordae - - Age and sex matched population 1

2

3

4

5

6

7

8

9

80

...>

70

rn

60

'0

50

::J

50

:::

o

1110)

..-

;>,

40

:0

30

...0 a.

20

Cll .0

... Posterior leaflet group } p=0.08 - - Anterior leaflet group

10

10 11 12 13 14 15

2

Postoperative years Fig. 1. Percent probability of survival of patients dismissed from hospital after repair of ruptured chordae tendineae of the mitral valve.

causes of the deaths of 10 patients were unknown. The probability of late survival among patients dismissed from the hospital was 92% ± 3% at 5 years, 73% ± 4% at 10 years, and 48% ± 6% at 15 years (Fig. I). Univariate analysis of multiple clinical variables and operative findings yielded only two factors that predicted poor late survival: age < 57 years (p = 0.02) and additional operative procedures at repair (p = 0.04). Rupture of chordae to the anterior leaflet, presence of atrial fibrillation preoperatively, and murmur of mitral regurgitation at dismissal. after valve repair were not significant risk factors for late death. With multivariate analysis, the only significant variable predictive of late death was the need for additional procedures at valve repair. The 5 year survival rate for patients with repair of ruptured chordae to the posterior leaflet was 90% ± 3%, compared with 69% ± 7% for patients with repair of chordae to the anterior leaflet (Fig. 2, upper panel). This difference in survival persisted throughout the 10 year follow-up period but was only suggestively significant (p = 0.08). When survival for patients who did not have associated procedures was analyzed, there was no difference in survival (p > 0.7, Fig. 2, lower panel). Twenty-eight patients required reoperation during the follow-up period (Fig. 3): 27 patients underwent mitral valve replacement and one had re-repair. Of the 28 patients, 20 had initial leaflet plication without resection, and they later experienced progressively severe mitral regurgitation. In each patient, the initial plication repair was intact; recurrent mitral insufficiency resulted from deterioration of the annuloplasty (10 patients), leaflet prolapse (six patients), and additional ruptured chordae (two patients). In two patients, the cause of failure of the valvuloplasty could not be determined. In

3

4

5

6

7

8

9

10

Postoperative years ~

;§!. 100

iii

90

> .s;

80

-

70 60

o

50

-

40 30

5

rn

;>,

:0

Cll

-g

a::

20 10

(50) (24)

(81

(9)

... Posterior leaflet group } __ Anterior leaflet group p>0.7

2

3

4

5

6

7

8

9

10

Postoperative years Fig. 2. Percent probability of late survival. Upper panel. After repair of ruptured chordae, comparing anterior with posterior leaflet rupture (p = 0.08). Lower panel. After repair of ruptured chordae, excluding patients undergoing simultaneous associated cardiac procedures (p > 0.7).

the remammg eight patients who had reoperation, symptomatic coronary artery disease (two patients), associated valvular heart disease (one patient), or failure of an alternate method of valve repair (five patients) were indications for reoperation, In the group of patients who had plication repair of ruptured chordae tendineae, the cumulative probability of reoperation because of progressivemitral insufficiency was 9% ± 3% at 5 years and 19% ± 5% at 10 years postoperatively (Fig. 4). Of the 27 patients subsequently requiring mitral valve replacement, 10 (36%) had previous repair of ruptured chordae to the anterior leaflet and 17 patients (64%) had prior repair of ruptured chordae to the posterior leaflet. No variables were predictive of the need for subsequent mitral valve operation. Importantly, there was no statistically significant difference in the risk of late reoperation between patients with repair of ruptured chordae to the anterior leaflet and those with repair of

Volume 89

Mitral regurgitation

Number 4

495

April. 1985

8 UJ

C Q)

toa.

~

7

Qj

3

E ::J

2

100

....CIl ...

90

0

5 4

~

c:

6

'0 .0

~

70

0

60

...

-.... Q)

0

>.

Z

.0

1-2

3-4

5-6

7-8

9-10

11-12 13-14

Postoperative years

Fig. 3. Reoperations from all causes after repair of ruptured chordae to mitral valve.

ruptured chordae to the posterior leaflet. Advanced age or presence of mitral valve prolapse, associated cardiac disease, or murmur of mitral regurgitation at dismissal after repair were not associated with a higher risk of subsequent reoperation for mitral regurgitation. At our institution during the same time interval under review, 106 patients underwent mitral valve replacement for mitral regurgitation resulting from ruptured chordae tendineae (Table I). In seven of these patients, initial valve repair was unsatisfactory and a prosthetic valve was inserted at the same operation. Follow-up was complete in 96% of patients. The two patient groups are different in that the mean age was 57 years for patients undergoing repair and 64 years for patients undergoing replacement (Table I). Important similarities, however, were noted. Operative mortality was 6.1% for the repair group and 7.5% for the replacement group. An advantage in late survival was seen in the repair group, in which the actuarial survival rate at 5 years was 92% compared with 72% after replacement (p < 0.003). Importantly, the incidence of late thromboembolism was 1.8 episodes/Itx) patient-years after repair and 8.0 episodes/f Oa patient-years after replacement.

Discussion Repairs of mitral regurgitation were among the first successful applications of open-heart surgery for acquired heart disease." Early techniques for replacing ruptured chordae with pericardium or synthetic material and methods of restoring mitral valve competence by bolstering the unsupported posterior leaflet with Ivalon sponge" gave variable results. A technique for plication of the unsupported segment of the leaflet was described in 1960. 7 With or without minor modifications, this method for repair of ruptured chordae tendineae has provided good results, as seen from our report and from

80

Q)

a.

All causes Plication (28)

50 40 30

CIl

20

... n,

10

.0 0

.......

(7)

(5)

(5)

(3)

1

2

3

4

5

6

7

8

9

10 11 12 13 14

Postoperative years

Fig. 4. Percent probability of reoperation. Probability of reoperation after repair of ruptured chordae tendineae for all causes (28 patients). Probability of reoperation for symptomatic recurrent mitral regurgitation in patientshaving plication repair (20 patients).

the experience of others. 18. 19 Triangular or quadrangular excisions of the unsupported segment of the involved leaflet have been proposed. 10, 1: Excision of the involved segment may allow easier approximation of the adjacent supported leaflet edges, but the practical advantages of this are debatable. With quadrangular excision, a portion of the base of the leaflet is also excised, and the reapproximated edges of the base of the valve reduce the circumference of the valve anulus. The proposed advantage of this maneuver is probably outweighed by the reduction of the anulus circumference afforded by a standard annuloplasty; leaflet dehiscence with this technique has occurred. 20 Dilatation of the mitral anulus is seen so frequently with mitral regurgitation secondary to ruptured chordae tendineae that we believe annuloplasty should be considered in every patient having valve repair. Preoperatively, compensatory mechanisms of cardiac hypertrophy and dilatation tend to maintain cardiac output by increasing left ventricular stroke volume. Cardiac decompensation may occur if associated mitral annular dilatation adds to the incompetence of the mitral valve mechanism. In such patients, leaflet plication alone would not completely restore valve competence. With ruptured chordae tendineae and annular dilatation, valve surface area may already be less than the annular orifice area, such that plication without annuloplasty would further aggravate this disproportion. Even with little or no annular dilatation, mitral valve annuloplasty may be advantageous after plication repair for ruptured chordae. Reduction of the circumference of the mitral valve anulus might

The Journal of

496

Orszulak et al.

reduce the stress imposed on the repaired leaflet and thus minimize the chance of disruption of the repair. Theoretical benefits of the addition of mitral valve annuloplasty to repair of ruptured chordae tendineae were appreciated early in this series, and this realization accounted for the use of annuloplasty in 88% of patients. Various techniques of annuloplasty have been applied,":" but the few patients in each category preclude any analysis of late results stratified by operative method. Our current preference is between the Kay21 and the Wooler" annuloplasty, modified by placing mattress sutures over felt pledgers," and the ring annuloplasty described by Carpentier and associates." Our report confirms the observations previously made that the operative risk for repair of ruptured chordae tendineae is low. The small number of deaths precludes any statistical analysis of risk factors other than the observation that the year of operation was an important determinant of mortality. The operative risk was 20% before 1966 and 1% from 1966 through 1980. Most deaths occurred before satisfactory valve prostheses were available. Two of the patients who died within 30 days of operation had additional mitral valve lesions, and two patients had aortic insufficiency; in retrospect, prosthetic replacement rather than valve repair would have been preferred. General improvements in technique of cardiopulmonary bypass, appreciation of the hazard of air embolism and-its prevention, and current techniques of myocardial protection may all have contributed to the decreased operative mortality in the latter part of this experience. In addition, it should be noted that an initial attempt at repair of ruptured chordae does not prejudice patient survival if the repair is unsatisfactory and valve replacement is necessary at the time of the original procedure. Among patients who underwent mitral valve replacement for ruptured chordae tendineae during this same time period, seven had an initial unsuccessful attempt at valve repair. There were no hospital deaths among these seven patients, and there were no complications that could be attributed to the additional time required for insertion of a prosthesis after unsuccessful valvuloplasty. A major potential shortcoming of repair of ruptured chordae is the impermanence of valvuloplasty and its influence on long-term survival. Our report adds important information on the durability of repair of ruptured chordae tendineae. Among patients who had valve repair with the plication technique, the cumulative risk of reoperation owing to persistence or recurrence of mitral insufficiency was low (9% at 5 years and 19% at

Thoracic and Cardiovascular Surgery

10 years). A review of results of repair of other mitral valve lesions indicated that most reoperations were necessary within the first several years after initial repair. This timing suggests that the causes were related to technical imperfections. 12. IJ Our results differ in that there was a slight but steady increase in the cumulative risk of reoperation throughout the follow-up period. This is consistent with gradual deterioration and may be attributed to progression of intrinsic mitral valve disease. In each patient who had prior plication, the valve at reoperation was found to have an intact leaflet repair, and valve insufficiency was due to annular dilatation or additional ruptured chordae. These findings suggest that late results might be improved by the uniform application of annuloplasty and by more secure methods of annular reduction. A major thrust of our investigation was to identify pathological and anatomic features that would predict a favorable result after repair of ruptured chordae tendineae. Considering operative risk, late survival, and need for subsequent reoperation, we found that the most serious potential precursor of a poor outcome was the presence of associated cardiac disease, specifically,associated aortic or mitral valve damage caused by endocarditis. As previously stated, many of these patients were operated on before reliable prosthetic valves were available, and the severity of associated valve abnormalities would currently mandate valve replacement. Others have adopted the technique of repair of ruptured chordae to the posterior (mural) leaflet but have cautioned against similar methods of repair of the anterior leaflet. IS, 19,25 The present study indicates that, in many cases, repair of ruptured chordae tendineae of the anterior leaflet can be managed satisfactorily by plastic reconstruction. Repair of ruptured chordae to the anterior leaflet may demand more precision during plication and careful testing to ensure an optimal hemodynamic result. Compared with plication of the posterior leaflet, in which a generous area of the leaflet can be imbricated, plication of the anterior leaflet with "overcorrection" can compromise the natural redundancy of the valve leaflet and produce commissural insufficiency. We found that valves made incompetent by ruptured chordae tendineae associated with mitral valve prolapse were amenable to repair, and our results were similar to those reported by Yacoub and associates. 13 In contrast to that of Kerth and colleagues," our experience supports the contention that careful closure of associated mitral valve clefts is indicated at the time of plication repair. Despite the detailed review of operative records from patients with ruptured chordae tendineae who under-

Volume 89

Mitral regurgitation 4 9 7

Number 4 April, 1985

went mitral valve replacement, we were unable to formally categorize clinical and pathological features that were contraindications to valvuloplasty; definite anatomic and hemodynamic factors that would outweigh surgical judgment regarding repair versus replacement could not be identified. In general, we believe that the quality of the tissues influences the suitability for repair and that leaflet or annular calcification or severe fibroelastic deficiency (thin, translucent leaflets) is an impediment to accurate, secure repair of ruptured chordae tendineae. We agree with Carpentier and associates" that the suitability of repair of ruptured chordae to the anterior leaflets is influenced by the amount of unsupported free edge of the leaflet. Carpentier and colleagues" consider valve repair contraindicated when resection (or plication) of more than one fourth of the leaflet is required. We have been hesitant to repair centrally located ruptured chordae of the anterior leaflet when adjacent chordae are extremely thin and elongated. Only one patient in our series underwent chordal shortening. Despite the potential pitfalls in analyzing nonrandomized groups of patients, the attempt seems justified to compare the results from repair of ruptured chordae tendineae with the results of valve replacement for ruptured chordae. The potential advantage of mitral valve replacement is the high degree of predictability with which a satisfactory hemodynamic result can be obtained. Fifty-eight percent of our patients who had repair of ruptured chordae had an apical systolic murmur detected at dismissal.These murmurs indicated some residual mitral insufficiency, although the murmurs typically either remained stable or progressed slowly. It should be noted that the presence of a murmur was not predictiveof a poor late functional result or need for reoperation. The durability of valve repair compares favorably with the intrinsic durability of prosthetic valves. This is especially true for replacement with a porcine valve prosthesis, in which primary tissue failure occurs in approximately 30% of patients by 9 years postoperatively." In our experience, the major advantages of repair of ruptured chordae tendineae are the excellentlate patient survival and the fourfold reduction of risk of late thromboembolism, when comparison is made with patients who have had valve replacement. Our results further suggest that, in borderline situations, the surgeon should accept a less-than-perfect anatomic result with repair rather than seek an ideal hemodynamic result with a valve prosthesis. We wish to acknowledge the assistance provided by Ms. Judith E. Kieffer and Ms. Donna M. Stucky in the secretarial

preparation of this manuscript. In addition, we are grateful for the assistance provided by the Department of Medical Statistics and Epidemiology and the Section of Medical Graphics in the preparation and presentation of this material. REFERENCES Selzer A: Nonrheumatic mitral regurgitation. Mod Concepts Cardiovasc Dis 48:25-30, 1979 2 Osmundson PJ, Callahan JA, Edwards JE: Ruptured mitral chordae tendineae. Circulation 23:42-54, 1961 3 Roberts WC, Braunwald E, Morrow AG: Acute severe mitral regurgitation secondary to ruptured chordae tendineae. Clinical, hemodynamic, and pathologic considerations. Circulation 33:58-70, 1966 4 Lillehei CW, Gott VL, De Wall RA, Varco RL: Surgical correction of pure mitral insufficiency by annuloplasty under direct vision. J Lancet 77:446-449, 1957 5 Scott HW Jr, Daniel RA Jr, Adams JE, Schull LG: Surgical correction of mitral insufficiency under direct vision. Report of clinical cases. Ann Surg 147:625-635, 1958 6 Guidry LD, Callahan JA, Marshall HW, Ellis FH Jr: The surgical treatment of mitral insufficiency by mitral annuloplasty. Mayo Clin Proc 33:523-531, 1958 7 McGoon DC: Repair of mitral insufficiency due to ruptured chordae tendineae. .J THORAC CARDIOVASC SURG 39:357-362, 1960 8 Ellis FH Jr, Frye RL, McGoon DC: Results of reconstructive operations for mitral insufficiency due to ruptured chordae tendineae. Surgery 59: 165-172, 1966 9 Kerth WJ, Sharma G, Hill JD, Gerbode F: A comparison of the late results of replacement and of reconstructive procedures for acquired mitral valve disease. J THORAC CARDIOVASC SURG 61:14-22,1971 10 Manhas DR, Hessel EA, Winterscheid LC, Dillard DH, Merendino KA: Repair of mitral incompetence secondary to ruptured chordae tendineae. Circulation 43:688-697, 1971 11 Selzer A, Kelly JJ Jr, Kerth WJ, Gerbode F: Immediate and long-range results of valvuloplasty for mitral regurgitation due to ruptured chordae tendineae. Circulation 45:Supp1 1:52-56, 1972 12 Carpentier A, Chauvaud S, Fabiani IN, De10che A, Reiland J, Lessana A, d'Allaines C, Blondeau P, Piwnica A, Dubost C: Reconstructive surgery of mitral valve incompetence. Ten-year appraisal. J THORAC CARDIOVASC SURG 79:338-348, 1980 13 Yacoub M, Halim M, Radley-Smith R, McKay R, Nijveld A, Towers M: Surgical treatment of mitral regurgitation caused by floppy valves. Repair versus replacement. Circulation 64:Suppl 2:210-216, 1981 14 Weldon CS, Krause AH, Parker BM, Clark RE, Roper CL: Clinical recognition and surgical management of acute disruption of the mitral valve. Ann Surg 175: 10001016, 1972 15 Sanders CA, Austen WG, Harthorne JW, Dinsmore RE,

The Journal of

498

16

17

18

19

20

21

22

23

24

25 26

Orszulak et af.

Scannell JG: Diagnosis and surgical treatment of mitral regurgitation secondary to ruptured chordae tendineae. N Engl J Med 276:943-949, 1967 Brandenburg RO: No more routine catheterization for valvular heart disease? N Engl J Med 305: 1277-1278, 1981 Kay EB, Nogueira C, Head LR, Coenen JP, Zimmerman HA: Surgical treatment of mitral insufficiency. J THoRAc SURG 36:677-696, 1958 Ross BA, Fox C, Brown AH, Webb-Peploe M, Braimbridge MV: Late results of valvoplasty for mitral regurgitation due to rupture of chordae of the posterior (mural) cusp. J THORAC CARDIOVASC SURG 71:533-536, 1976 Messmer BJ, Gattiker K, Rothlin M, Senning A: Reconstruction of the mitral valve. Ann Thorac Surg 16:30-43, 1973 Carpentier A, Reiland J, Deloche A, Fabiani J-N, D'Allaines C, Blondeau P, Piwnica A, Chauvaud S, Dubost C: Conservative management of the prolapsed mitral valve. Ann Thorac Surg 26:294-302, 1978 Kay JH, Egerton WS: The repair of mitral insufficiency associated with ruptured chordae tendineae. Ann Surg 157:351-360,1963 Wooler GH, Nixon PGF, Grimshaw VA, Watson DA: Experiences with the repair of the mitral valve in mitral incompetence. Thorax 17:49-57, 1962 Danielson GK: Surgical treatment of acquired heart disease, Cardiac and Vascular Diseases, Vol 2, HL Conn Jr, 0 Horwitz, eds., Philadelphia, 1971, Lea & Febiger, pp 1365-1414 Carpentier A, Deloche A,'Dauptain J, Soyer R, Blondeau P, Piwnica A, Dubost C: A new reconstructive operation for correction of mitral and tricuspid insufficiency. J THORAC CARDIOVASC SURG 61:1-13,1971 West PN, Weldon CS: Reconstructive valve surgery. Ann Thorac Surg 25: 167-177, 1978 Gallo 1, Ruiz B, Nistal F, Duran CMG: Degeneration in porcine bioprosthetic cardiac valves. Incidence of primary tissue failures among 938 bioprostheses at risk. Am J Cardiol 53: 1061-1065, 1984

Discussion DR. JOHN RELLAND Paris. France

We used the plication technique at the Broussais Hospital in Paris in the early days of cardiac surgery. However, we found that better results could be obtained in ruptured chordae of the posterior leaflet by quadrangular resection and subsequent remodeling of the anulus with a Carpentier prosthetic ring. The prosthetic ring is always used to reinforce this repair and to prevent recurrent dilatation, which is probably responsible for the high reoperation rate that the authors reported. In ruptured chordae of the anterior leaflet, we found, as the authors did in their experience, that plication or resection was associated with a higher incidence of either residual or

Thoracic and Cardiovascular Surgery

recurrent regurgitation. We now prefer to use transposition of the chordae from the posterior leaflet to the anterior leaflet, as presented at this meeting for the first time last year by Dr. Carpentier. One year later, we are in a position to confirm that these preliminary results using this technique have held up with time. Our current results using these techniques in this particular group of patients show a reoperation rate of 0.6%/patient-year. DR. JEROME H. KAY Los Angeles. Calif

We operated on 325 patients from 1959 to 1984 with significant mitral regurgitation. Twenty-seven (11%) had stretched chordae tendineae and 127 (54%) had torn chordae tendineae. We were able to repair the valve in 235 patients but had to replace the mitral valve in 90 patients. Our method of repair is the same one that we have used since 1959 and differs somewhat from that of the authors. If there are areas of torn or stretched chordae tendineae, that portion of the leaflet is sutured to the nearer papillary muscle. This is true for the anterior and the mural leaflets. All patients with mitral regurgitation have a dilated mitral anulus. It is also important to do away with two thirds of the mural anulus to decrease the mitral orifice and correct the insufficiency. Whether the mitral valve was replaced with a bioprosthesis or a disc valve, the survival probability was the same. However, we compared the 90 valve replacement patients with the 235 valve repair patients and found a significant difference in survival probability at 10 years. The chance of survival with repair was 70% ± 4% compared to 42% ± 7% with replacement. The p value was 0.001. The incidence of embolism is small with repair compared to replacement. There were 14 peripheral or cerebral embolic events in 1,520 patient-years (0.92%). With mitral repair for mitral regurgitation, the patient requires anticoagulant therapy for the first 3 months only followingoperation. The need for reoperation for the entire group of patients with repair for the past 26 years was 12%. DR.ORSZULAK(C0smm I would like to thank the discussers for their comments. We do not have enough experience with quadrangular resection to make any comparative statements about plication versus the quadrangular resection. However, we are concerned about the isolated reports for this area dehiscing in late follow-up. Dr. Kay, we agree that the mural leaflet is the area to make up the difference in circumference involvingannuloplasty. Our incidence of 1.8 episodes per 100 patient-years in the repair group was relatively high, but this was early in our experience and many of those patients, despite repair, did not have even short-term anticoagulation. However, at present we do believe that short-term anticoagulation is recommended even in the repair groups.