- Email: [email protected]
Results of mitral valvuloplasty with a suture plication technique
J
THORAC CARDIOVASC SURG
79:349-357, 1980
Results of mitral valvuloplasty with a suture plication technique Between January, 1975, and October, 1978, a total of 243 patients underwent repair of the mitral valve with a suture plication technique. Mitral valve replacement (MVR) was performed in 36 cases in which significant residual regurgitation was apparent on testing the valve after repair. The hospital mortality rate was 5.7%. Five patients required MVR within I month of operation. A total of 190 patients were discharged from hospital with what was considered to be a satisfactorily functioning mitral valve. Excluding patients from overseas, detailed follow-up information is available in 80 cases. Of these 80 patients, 13 have subsequently undergone MVR. Factors favorably affecting survival and durability of repair are degenerative valve disease and age below 55 years. Clinical and echocardiographic assessment indicate that this method of repair initially produces good symptomatic improvement and a normal or nearly normal pattern of left ventricular filling. The good early results are only maintained in patients with degenerative valve disease. We therefore no longer use or recommend this technique for elderly patients with rheumatic valve disease.
D. F. Shore, M.D. * (by invitation), P. Wong, M.D. (by invitation), and M. Paneth, M.D. (by invitation), London, England Sponsored by Mortimer J. Buckley, M.D., Boston, Mass.
In 1977 we 1 reported the early results of our first year's experience with mitral valve repair with the mitral plication suture (MPS). The MPS was used either to reduce the mitral anulus to an appropriate size or to correct regurgitation confined to the anterolateral or posteromedial half of the valve, which either preexisted operation or occurred following adequate relief of valve stenosis. The MPS is a double suture of 2-0 Ethiflex guarded with a Teflon pledget. It is placed at the margin of the central fibrous body and sutured circumferentially around the anulus in 2 to 3 mm steps to the posterior portion of the anulus, where it is tied over a second pledget (Fig. I). A second suture is placed at the opposite side if necessary. In cases of ruptured chordae of the posterior leaflet, the plication suture is continuous from one margin of the central fibrous body to the other, so that the leaflet repair will be supported posteriorly. From the Brompton Hospital, London, England. 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: M. Paneth, The Brompton Hospital, London, England SW3. 'Present address: The Hospital of the Albert Einstein College of Medicine, 1825 Eastchester Rd., Bronx, N. Y. 10461.
In an earlier report, we emphasized that the MPS was only part of an aggressive systematic approach to the conservation of the mitral valve. Success depended on each subunit of the mitral valve being individually scrutinized and specifically treated. Early follow-up (average 10 months) indicated that good clinical results could be obtained with a low operative mortality rate. In addition, the overall rate and pattern of left ventricular filling determined by echocardiography were normal or nearly normal in all patients studied. 1 Present series
Encouraged by these early results, we attempted repair of the mitral valve in 243 patients between January, 1975, and October, 1978. All patients were in Functional Class III or IV preoperatively. The total number of operations performed on the mitral valve during this same period was 423. The operation was performed in all cases through a median sternotomy. Cardiopulmonary bypass was established by standard aortic and bicaval cannulations. Initially, the body temperature was reduced to 32° C and the myocardial temperature was reduced to 15° C by selective aortic root perfusion. During the last 12 months of this study we routinely employed cardiople-
0022-5223/80/030349+09$00.90/0 © 1980 The C. V. Mosby Co.
349
The Journal of
350 Shore, Wong, Paneth
Thoracic and Cardiovascular Surgery
Table I. Associated procedures Procedure
No.
MR alone MR + AVR MR + tricuspid repairs MR + A VR + tricuspid valve repair MR + tricuspid valve repair MR + secundum ASD MR + primium ASD MR + CAVG MR + aortic valve repair
123 36 24* 14 I I I 2
3
Legend: MR, Mitral repair. AVR, Aortic valve replacement. ASD, Atrial
septal defect. CAVa, Coronary artery vein graft. 'Including one secundum atrial septal defect and one coronary artery vein graft.
Table II. Hospital deaths Cause
No.
Low cardiac output Preoperative systemic embolus Development of left atrial thrombus Secondary hemorrhage Empyemalsepticemia Uncertain
7 I I I I 3
Table III. Causes of late failure of repair Cause Mitral regurgitation Complete dehiscence of MPS Partial dehiscence following infection Partial dehiscence in the absence of infection Stretching of remaining chordae tendineae Progressive disease of subvalvular apparatus Shrinkage of valve leaflets Restenosis Hemolysis
No.
10 I I 3 I 2 2
the prolapsing area or by shortening of individual chordae." Adherent chordae were separated by sharp dissection down to and including papillary muscles. One or two MPSs were then inserted as described previously. The repair was tested by rendering the aortic valve incompetent and allowing the ventricle to fill. We have found that this maneuver produces a sufficient increase in left ventricular pressure to close the mitral valve and demonstrate the presence of any residual incompetence. When bypass was discontinued, simultaneous left atrial and left ventricular pressures were measured to exclude the presence of a significant transmitral gradient. Additional procedures were performed as indicated in Table I. Statistical evaluation. Three-year survival figures were calculated by the actuarial method." Ninety-five percent confidence limits were determined by the Greenwood formula. 4 Assessment of valve function after repair. We have found echocardiographic assessment of left ventricular filling to be a reliable means of evaluating mitral valve function after repair. Real-time M-mode echocardiograms were recorded, digitized, and computer analyzed." Linear traces of left ventricular dimension and the rate of change of cavity size were derived. The peak rate of dimension change (PRDC) of the transverse dimension of the left ventricle reflects the peak volume of blood flow through the mitral valve. Thus the PRDC can indicate whether the valve orifice is normal (10 to 20 cm/sec), stenotic « 10 cm/ sec), or regurgitant (>20 ern/sec). Results
3 I
Legend: MPS, Mitral plication suture.
gic arrest together with topical cooling of the myocardium and reduction of body temperature to 28° C. The left atrium was opened posterior to the interatrial groove and the mitral valve was inspected to assess its suitability for repair. Valves with rigid, calcified orifices, with large amounts of calcium in the chordae tendineae or papillary muscles, or with destruction of more than half of the tissue of the anterior leaflet were not considered suitable for repair. In all other cases repair was attempted. Abnormal commissures were split along their line of fusion to within 2 mm of the anulus. Leaflets were decalcified and valve clefts or perforations were sutured with interrupted sutures of 4-0 Ethiflex. Prolapsing leaflet margins were repaired by trapezoidal excision of
Hospital deaths. There were 14 deaths (5.7%) within I month of operation. The causes of death are described in Table II. All seven of the patients dying with low-output syndrome were considered to have a properly functioning valve. Two deaths followed emergency mitral valve replacement (MVR) in the early postoperative period. Operative failures. Of the 246 attempted repairs, 36 were judged to be unsatisfactory and the valve was immediately excised and replaced with a Bjork-Shiley prosthesis (Pyrolite disc size 29 or 31). Ninety percent of the operative failures occurred in patients with advanced rheumatic valve disease. Failure resulted either because it was impossible adequately to free the individual components of the subvalvular apparatus or because of significant residual regurgitation after completion of the repair. Early failure of repair. Careful clinical evaluation of mitral valve function was made postoperatively. In those cases in which there was any doubt that function
Volume 79
Mitral valvuloplasty
Number 3 March,1980
Central fibrous body
,
Anterior leaflet
'"\
/\
I I
100 (80) ..... (80)''............ 80
\
v» a.::
\ \
> > a.::
I I / / /
------(19) ......
---(19)
o 60
\ \
II
'- -
35 1
~
40
ltI!
Posterior leaflet
Fig. 1. Mitral plication suture inserted on both sides of the mitral valv~.
was satisfactory, echocardiographic assessment of left ventricular filling aided in the evaluation. Five patients required MVR in the early post-operative period, three owing to mitral regurgitation following partial dehiscence of the repair and two owing to inflow obstruction. The latter probably was not recognized at operation because the cardiac output was too low to produce a significant gradient. One hundred ninety patients were discharged from the hospital and were considered to have a satisfactorily functioning mitral valve. Long-term follow-up. Of the 190 patients discharged, 110 were from overseas and therefore were not available for detailed clinical and hemodynamic assessment. This study therefore concerns the followup of the remaining 80 patients. The average age of these 80 patients at operation was 54 years, contrasting sharply with the average age of 33 years of those patients from overseas who were not included. Of the 80 patients, 13 have subsequently required MVR. The causes of failure are listed in Table III. Seven patients in this series required MVR because of progression of rheumatic valvular disease. The patient who required reoperation for hemolysis had a very small regurgitant jet. This was readily corrected without the need for valve replacement and there was no evidence of hemolysis postoperatively. There have been nine late deaths, four following MVR, one following a cerebrovascular accident, and four of unknown cause. The overall survival rate, calculated by the actuarial method, is 89% at 3 years (Fig. 2). The chance of surviving without requiring reoperation is 72% at 3 years (Fig. 2).
20
- - % Survival
--- % Survival with Natural Valve
Fig. 2. Survival 3 years following mitral repair.
100
--
.....
80
, .....
I
',----- I
Vl
a.::
0 60
> > a.::
::::>
Vl
ltI!
±2SD 40 20
Age
-<50 ---)50
2 YEARS Fig. 3. Influence of age on survival following mitral repair.
The influences of age (Fig. 3), surgical pathology (Table IV, Fig. 4), and associated valve disease (Fig. 5) on postoperative survival were evaluated statistically by comparing survival curves after the entire group of patients had been divided into appropriate groups. At 3 years, significantly adverse correlates of survival were age above 50 years and rheumatic valvular stenosis. Associated valve disease did not adversely affect survival. Patients with degenerative valve disease had a significantly improved chance of event-free survival when compared with patients with rheumatic valvular disease, either predominantly stenotic or predominantly regurgitant (Fig. 6). Incidence of thromboembolism. Patients were not given anticoagulants unless aortic or tricuspid valve
The Journal of Thoracic andCardiovascular Surgery
352 Shore, Wong, Paneth
,,
.~------i-I "
80 V)
a.:
0
> > a.:
::::>
V)
60
",----1
±2 SD 40
~
20
w w 80 a.:
LL.
f--
zw
60 > w V)
a.:
Aetiology --- Rheumatic Incompetence _. _. Rheumatic Stenosis
2
.....
,,
,
......
-----_...:..~
,,
. !. I
"
±2SD
Aetiology ~ 40 > --- Rheumatic Incompetence a.: ::::>
V)
~
- - Degenerative
,,
20
_. _. Rheumatic Stenosis - - Degenerative
3
YEARS
Fig. 4. Influence of surgical pathology on survival following mitral repair.
80
-.. . . . . . . . . . -----.:1
V)
a.:
o
> > a.:
::::>
V)
60 ±2SD 40
- - - Multiple Valve Disease --Mitral Valve Disease Only
20
Fig. 6. Influence of surgical pathology on survival without the need for subsequent valve replacement following mitral repair. Table IV. Etiology
Rheumatic Degenerative Infection (rheumatic valve) Infection (normal valve) Coronary artery disease Congenital Uncertain
MS
MR
19
25 19 5 I
1 I 8
Legend: MS, Mitral stenosis. MR. Mitral repair.
Table V. Echocardiography (49 patients) PRDC Normal range (10-20 em/sec) Stenotic «10 em/sec) Regurgitant (>20 em/sec)
I No. of patients 34
12 3
Fig. 5. Influence of associated valvular disease on survival following mitral repair.
Legend: PRDC. Peak rate of dimension change.
replacement had been required. Definite evidence of thromboembolic disorders is available in two cases. One patient died following a cerebral embolus and a second patient developed a right hemiplegia and has residual weakness. Clinical and hemodynamic assessment. Of the 71 patients still alive, 62 have not required valve replacement. All but five of these patients were symptomatically improved following operation. Eight patients improved by three grades, 34 patients by two grades, and 15 patients by one grade. The condition of 49 patients, including four of the
five patients who were not symptomatically improved was assessed by echocardiography (Table V). Thirtyfour patients had a peak rate of dimension change (PRDC) within the normal range (Table V). Twelve patients had a PRDC below 10 em/sec, but only two of these patients had a PRDC below 7 em/sec. All the patients with a PRDC below 10 em/sec were in Functional Class I or II. The three patients who had a PRDC in the regurgitant range were in Functional Class II. Of the five patients who were not symptomatically improved following operation, all had a PRDC within normal range.
Volume 79 Number 3 March. 1980
Eleven of the 15 patients with a PRDC outside the normal range had rheumatic valve disease; four of these patients had predominantly stenotic lesions and seven, predominantly regurgitant valves. Seventeen patients with degenerative valve disease were examined by echocardiography. Only one of these patients had an abnormal PRDC of 23 ern/sec, indicating slight mitral regurgitation. Discussion
A durable and easily reproducible method of conserving the diseased mitral valve remains an attractive goal. The advantages of repair over prosthetic replacement are a low incidence of thromboembolic episodes without, in most cases, the need for anticoagulation, and the retention of a valve whose function is normal or more nearly normal than the best of available prostheses. However, to be acceptable, any method of repair must have an operative mortality rate and postoperative event-free survival rate which compare favorably with those following valve replacement with either mechanical prostheses'":" or tissue valves.?: 10 The operative mortality rate (5.9%) and 3 year survival rate (89%) in our series indicate that this method of repair satisfies these criteria. Repair with the MPS promised to be quick and simple to perform. If unsuccessful, the valve could then be readily excised and replaced without much loss of time. These factors, together with the fact that the valve ring retained the normal physiological ability to shorten its circumference during ventricular systole," led us to believe that the MPS technique would prove an advantage over methods of repair which depended on the use of annuloplasty rings. The high incidence of operative failures was due in part to the broad criteria for selection, which included patients with advanced rheumatic disease provided only that extensive calcification was absent. In addition, the technique we used for testing the valve may have overestimated the number of patients with significant residual regurgitation. In some of these cases the regurgitation may have been less severe had competence of the valve been assessed with a contracting ventricle, physiological shortening of the mitral ring circumference, and normal development of intraventricular pressure. However, one of the advantages of such overestimation of the number of failed repairs was that there were no hospital deaths in this group of 36 patients. A method of testing repair procedures that comes close to the ideal has been described by Nair and assoelates." An appraisal of the early results of this technique
Mitral valvuloplasty
353
suggested that predictably good symptomatic results could be obtained together with a normal or nearly normal rate and pattern of left ventricular filling. This is in contrast to the situation after MVR with a BjorkShiley or Starr-Edwards prosthesis, which produced both an abnormal rate (PRDC; Bjork-Shiley = 7 to 8 em/sec; Starr-Edwards = 6 ern/sec) and an abnormal pattern of left ventricular filling. However, the number of patients who have required MVR within 3 years of the initial operation is disappointing. We have been unable to match the durability of the repair that has been obtained with the use of annuloplasty rings. In addition, Duran and Ubago" compared left ventricular function following the use of a rigid ring and a flexible ring and found no significant difference. Therefore, the theoretical advantage of retaining ring flexibility may in practice be minimal. Although dehiscence was responsible for failure of the repair in four patients in the absence of infection, progression of the disease was responsible in a further eight patients. Seven of these eight patients had rheumatic valve disease. Progression of the disease is rarely cited as a cause of failure in other series. However, one difference, which may have had a significant influence in this respect, is the average age of 54 years of patients in our series compared with the average age of patients in the series reported by Duran and Ubago.P 37 years, and Carpentier and associates ,14. 15 40 and 43 years. As pointed out earlier, those patients from overseas whose average age was 33 years could not be included in the long-term follow-up. However, as a result of the high incidence of MVR following repair, we have abandoned attempts to conserve the mitral valve in the elderly patient with rheumatic valve disease except in those cases in which only a commissurotomy is required. On the other hand, repair of the degenerative valve has been followed by a 94% survival at 3 years and a low incidence of subsequent MVR. All 19 of the patients in this study with degenerative valve disease had ruptured chordae; 18 patients had rupture of chordae to the posterior leaflet and one patient, rupture of chordae to the anterior leaflet. In addition, there was redundancy of that part of the affected leaflet which had lost its chordal suppost. Slight-tosevere mitral ring dilatation was also present. It is our practice in these cases to perform a trapezoidal excision of the redundant leaflet tissue. After excision, the cut edges of the leaflet are approximated by interrupted sutures of 4-0 Ethiflex. The MPS is inserted to narrow the ring and support the leaflet repair. In our experience progression of the disease is rare in this group; stretching of the remaining chordae was
The Journal of Thoracic and Cardiovascular Surgery
354 Discussion
responsible for recurrent regurgitation necessitating MVR in only one case. The symptomatic results and the hemodynamic status of this group of patients is excellent following operation. The results are clearly superior to those obtained following repair of rheumatic valves, because leaflet pliability and the function of the subvalvular apparatus are normal following repair. Conclusion
The operative mortality rate (5.9%) and 3 year survival rate (89%) after repair of the mitral valve with a suture plication technique compare favorably with those after MVR. The restoration of normal mitral valve function after repair, demonstrated by echocardiography in the early postoperative period, is not maintained in patients with rheumatic valvular disease. Deterioration of function is more common in patients over 55 years of age. This has resulted in 16% of patients requiring MVR within 3 years. We therefore no longer use or recommend the use of this technique for repair of the valve in the elderly patient with rheumatic valve disease. Good functional results are maintained in patients with degenerative valve disease. The use of this technique in younger patients with rheumatic disease requires further evaluation.
2
3 4
5
6
7
8
REFERENCES Burr HB, Krayenbuhl C, Sutton MStJ, Paneth M: The mitral plication suture. J THORAC CARDIOVASC SURG 73:589-595, 1977 McGoon DC: Repair of mitral insufficiency due to ruptured chordae tendineae. J THoRAc SURG 39:357-362, 1960 Grunkemeier GL, Starr A: Actuarial analysis of surgical results. Ann Thorac Surg 24:404-408, 1977 Cutler SJ, Ederer F: Maximum utilization of the life table method in analyzing survival. J Chron Dis 8:669-672, 1958 Gibson D, Brown DJ: Measurement of instantaneous left ventricular dimension and filling rate in man using echocardiography. Br Heart J 35:1141-1149,1973 Starr A: Mitral valve replacement with ball valve prosthesis: A current appraisal of late results, The Mitral Valve, D Kalmanson, ed., Acton, Mass., 1976, Publishing Science, Inc., Publisher, p 413-429 Bjork VO: Results of mitral valve replacement with the Bjork-Shiley tilting disc valves, The Mitral Valve, D Kalmanson, ed., Acton, Mass., 1976, Publishing Science, Inc., Publisher, p 449-455 Sutton MStJ, Miller GAH, OIdershaw P, Paneth M: Valve replacement with the Bjork-Shiley valve. Experience with 390 patients. Read at the Summer Meeting, The
Thoracic Society, Liverpool, England, July 10, 1976 9 Ionescu MI, Tandon AP, Mary DAS, Abid A: Heart valve replacement with the Ionescu-Shiley pericardial xenograft. J THoRAc CARDIOVASC SURG 73:31-42, 1977 10 Angell WW: Long term results of tissue valve grafts for mitral replacement, The Mitral Valve, D Kalmanson, ed., Acton, Mass., 1976, Publishing Science, Inc., Publisher, p 481-488 II Tsakiris MD: The physiology of the mitral annulus, The Mitral Valve, D Kalmanson, ed., Acton, Mass., 1976, Publishing Science, Inc., Publisher, p 21-26 12 Nair KK, Yates AK: Direct evaluation of mitral valve function during surgery following conservative procedures. J THORAC CARDIOVASC SURG 73:684-685, 1977 13 Duran CG, Ubago JLM: Clinical and hemodynamic performance of a totally flexible ring for atrioventricular valve reconstruction. Ann Thorac Surg 22:458-463,1976 14 Carpentier A, Reiland J, Deloche A, Fabiani J, D'AIlaines C, Bordeau P, Piwnica H, Chauvaud S, Dubost C: Conservative management of the prolapsed mitral valve. Ann Thorac Surg 26:294-302, 1978 15 Carpentier A: Plastic and reconstructive mitral valve surgery, The Mitral Valve, D Kalmanson, ed., Acton, Mass., 1976, Publishing Science, Inc., Publisher, p 527-540
Discussion (Papers by Reed [page 321), Duran [page 326], Carpentier [page 338], Shore [page 349], and their colleagues)
DR. ROBERT W. M. FRATER Bronxville. N. Y.
It appears to me that there are three problems with plastic repairs of the mitral valve. The first is simple: "Life is short and the art is long" (Hippocrates). The surgeons who have talked today are clearly artists, and Dr. Carpentier, the doyen of this school of artists. They have spent the time and devoted the intense effort that is necessary for the development of artistic skill. Not all of us have done that, and therefore not all of us can remotely hope to achieve their results. The second problem is this: How can one be sure that the valve is neither stenotic nor insufficient at the end of the repair before closing the atrium? Drs. Reed and Carpentier particularly, and the others, too, have suggested that this may be possible using their artistic skill. Third, do the repairs really hold up with time? The evidence from the papers herein presented from four different countries is that some of the repairs certainly do seem to hold up with time. We do have the problem of comparing apples with oranges-young rheumatic patients with pure insufficiency, older rheumatic patients with mixed stenosis and insufficiency, and older patients with degenerative valvular disease. From what we have learned today, it is very clear that the young rheumatic patient with pure insufficiency does well. However, Dr. Shore's paper suggests very strongly that
Volume 79
Discussion
Number 3 March,1980
valvuloplasty in the elderly rheumatic patient, particularly the one with insufficiency and stenosis, does not hold up well, whereas valvuloplasty in the elderly patient with degenerative disease does. Regarding the second question-how to be sure that the right results have been achieved at the time of operation-I would like to suggest that one of the most difficult problems facing the surgeon at the operating table is to decide how to shorten elongated chordae so as to produce a competent valve. Dr. Carpentier has described a technique for this which he clearly executes very well. However, the art lies in deciding exactly how much the chordae should be shortened. There is a way of deciding this provided that only one cusp has elongated chordae. The rules are very simple: Both cusps hang to the same level in the ventricle, and the distance from a point on a papillary muscle to the true edge of either cusp is always the same and always less that the distance to the atrioventricular ring. Given two cusps, one with elongated chordae and one with chordae of normal length, if the free edge of the prolapsing cusp is stitched temporarily to the free edge of the nonprolapsing cusp and the cusps are gently pulled on to put the shorter group of chordae under slight tension, it then seems possible to perform Dr. Carpentier's maneuver knowing full well that the end result will be correct.
DR. JEROME HAROLD KAY Los Angeles, Calif.
I am so pleased that other surgeons are reporting good results with mitral repair for mitral insufficiency, since I have been advocating this for the past 20 years. The incidence of peripheral emboli is far lower with repair than with mitral valve replacement with the porcine valve. One hundred sixty-five patients had mitral repair for pure mitral insufficiency from I to 20 years ago. Fifteen patients were reoperated upon for recurrent mitral insufficiency or mitral stenosis. I am quite sure this figure is far lower than the number of patients requiring reoperation following mitral replacement with a porcine valve. Regarding the surgical technique, we have always used sutures alone, never mitral rings. I feel strongly that the use of a ring is unnecessary for the surgical treatment of pure mitral insufficiency. All patients with pure mitral insufficiency have a dilated anulus , and 60% of the patients also have tom chordae tendineae or ruptured papillary muscle. For repair, the leading edge of the valve at the area of the tom chordae tendineae is sutured down to the papillary muscle nearest to this area. If an anterior leaflet has the tom chordae tendineae, that part of the valve has to be sutured to the papillary muscle. If the mural leaflet has the tom chordae tendineae, the repair has to be performed so as to allow the anterior leaflet to balloon out and abut against the mural leaflet. This is extremely important in the technique. With tom chordae tendineae of the mural leaflet, the area of the leaflet with the tom chordae tendineae is sutured to the papillary muscle nearest the tear.
355
Seventeen percent of the patients have stretched chordae tendineae along with the dilated anulus, and 23% have a dilated anulus only. For the repair, the area of the mural leaflet with the stretched chordae tendineae should be sutured down to the nearest papillary muscle. One may have to suture the mural leaflet down to the papillary muscle in two or three different areas. If the anterior leaflet has stretched chordae tendineae, the anterolateral part of the anterior leaflet has to be sutured to the anterolateral papillary muscle to allow the anterior leaflet to balloon out. In all of the patients with pure mitral insufficiency, whether or not there are stretched chordae tendineae, there is always a dilated anulus. In repairing the dilated anulus, sutures are placed at the posteromedial and anterolateral commissural areas to do away with approximately 70% of the mural anulus. Care must be taken not to encroach on the anulus of the anterior leaflet. Replacement is far easier than repair in most patients; however, with experience, repair is feasible in approximately 75% of the patients and is far better for the patient than replacement. MR. MARK BRAIMBRIDGE Landon, England
As I see it, now that most of us are using cardioplegia, the problem in mitral valve repair is testing the valve. I would like to ask each of the discussers to tell us how they test their valves and what constitute their immediate criteria of successful repair. In our experience, when the ventricle is completely relaxed it is difficult to know how the tone in the valve ring will be when the heart is once again reperfused. We dilate the ventricle gently with cold saline through the left ventricular vent to assess the competence of the valve. This method is very empiric, but it does seem to give us a fair idea of how the valve will be after the operation. I would be most interested to know how the participants test the valves. DR. JOHN W. KIRKLIN Birmingham. Ala.
A question directed toward Dr. Carpentier has come to the desk from the audience. It is as follows: At reoperation upon a valve that has been repaired with a ring, is the valve now so small because of the ring that it is difficult to insert a prosthesis of adequate size? I would like to ask a question of each of the essayists: Is it a particular problem if a patient with mitral incompetence has ruptured chordae or very elongated chordae to the anterior leaflet? DR. R E ED (Closing) I was pleased to see that despite the variety of techniques used, the results were uniform. The operative mortality rate for this group may be explained partially by their youth and by the selectivity in excluding patients with heavy calcification. The very low incidence of thromboembolism is similar
The Journal of
35 6 Discussion
Thoracic and Cardiovascular Surgery
irrespective of technique and is considerably lower than that reported for valve replacement. It was good to learn of Dr. Kay's results in patients with coronary artery disease. I have followed his work with great interest and respect over the years and often have felt that he was the only other voice in the wilderness. I want to thank Dr. Frater for his kind remark, although I can take very little credit for being an artist. We have tried to describe a technique that has mathematical precision and is readily duplicated. It can be learned easily, but meticulous attention to detail is necessary to make it work. To answer Mr. Braimbridge's question: We deliberately make the valve smaller than is probably necessary but, by so doing, almost ensure that it will be competent. The valve is tested routinely by coming off bypass and elevating the pressure with a systemic pressor. This increase in peripheral resistance is necessary to assure that the valve is being appropriately stressed. During this time we palpate for regurgitation through an untied suture in the left atrial closure. Yes, Dr. Kirklin, the most troublesome patients are those with elongated chordae to a prolapsed anterior leaflet. This condition is treated by excising a wedge of tissue and then bringing together valve edges that are adequately supported by chordae which are not stretched. Most of this free edge no longer participates in the line of closure because the annuloplasty repair advances the major leaflet to form a new line of closure. All patients have annuloplasty regardless of whether they have had a valvuloplasty procedure. DR. 0 U RAN (Closing) Regarding the technical difficulties of performing this type of conservative operation, I would like you to think for a moment how difficult we would consider these procedures if mitral replacement had been developed before mitral commissurotomy. How many commissurotomies would be done today? Although in our unit the residents learn first to replace a valve and then to conserve it, it is not a matter of artistry but of craftmanship. A chronologie study of the percentage of valve replacements versus conservation, per individual surgeon of our unit, shows that it takes some time for the newly arrived to this technique to be convinced of its usefulness. A steady rise in the number of annuloplasties is seen with time. I would therefore be glad if you will just evaluate the next mitral valve before taking it out. Our method of checking valve competence intraoperatively consists of filling the left ventricle directly through the left vent, which is temporarily connected to the arterial line. It was published in The Annals a/Thoracic Surgery. One of our surgeons thought that, under cardioplegic conditions, we might be getting too high a rate of insufficiency. Back in the animal laboratory, we saw that the normal mitral valve in a flaccid heart did not regurgitate. Finally, I will address the question of more complex techniques, such as chordal shortening. We have performed only
18 such operations and, although very satisfied with this technique, I think the experience is too short for a comment. Such operations are difficult and take time. We therefore keep it simple with a straightforward annuloplasty after a good commissurotomy.
DR. CAR PEN TIE R (Closing) There have been many interesting questions. In answering Dr. Frater's question, we have demonstrated that the repair holds up with time. His suggestion of suturing the free edges of the leaflet to assess the degree of shortening is an interesting idea, and I congratulate him. Dr. Kay has been a pioneer in this field, and it is appropriate for him to remind us that elongated chordae occur more frequently than expected. However, I do not share his opinion that some sort of anulus remodelling is not necessary when the anulus is dilated. Mr. Braimbridge raised the question of the difficulty in assessing valve function when using cardioplegia. This is the reason why cardioplegia should be instituted only after valvular function has been evaluated and the various lesions have been analyzed carefully. Dr. Kirklin touched on the problem of the prolapsed anterior leaflet. If the cause of the prolapse is a chordal rupture, only rupture of one or two chordae can be treated by leaflet resection. Rupture of more than two chordae would require extensive resection of the anterior leaflet and therefore a lack of tissue. If the prolapsed leaflet is due to elongated chordae, the condition can be treated by chordal shortening. The degree of shortening is determined by measuring the degree of overriding of the free edge of the anterior leaflet above the plane of the orifice. Finally I would like to elaborate on two limiting factors which may explain why surgeons react differently when faced with the same lesion. The first is difficulty in patient selection. Patient selection depends on valvular lesions. Valvular lesions can be diagnosed prior to the operation, but most accurately during operation. It is critical for the surgeon to be able to recognize all of the lesions. Since there is only one technique for each lesion. the problem is not to decide which technique should be used but rather to recognize the lesion. The second point is reproducibility of the techniques. This depends upon the will of the surgeon to train himself in order to acquire the necessary experience. It also depends upon our ability to simplify the techniques. In this regard, and answering Dr. Duran's question, we have recently been able to simplify chordal shortening. The details of the modified technique are presented in the paper. Reproducibility of the techniques also depends on the quality of exposure. Good exposure is obtained by (1) lysis of any adhesions present, (2) extensive incision of the atrium, (3) placing a pad in the pericardium to expose the papillary muscles, and (4) using a self-retaining retractor.
Volume 79
Discussion
Number 3 March,1980
DR. S H 0 R E (Closing)
To answer Mr. Braimbridge's question, we tested the mitral valve after repair by rendering the aortic valve incompetent and allowing the ventricle to fill. We also took care to measure left ventricular and left atrial pressures simultaneously at the discontinuation of cardiopulmonary bypass to exclude the presence of a significant transmitral gradient. Five patients required valve replacement in the early postoperative period in the first year of our experience with this technique, but since then only one patient has required valve replacement in the early postoperative period. These results indicate that we are getting better at determing whether or not the valve repair has been adequate. To answer Dr. Kirklin's question about the management of
35 7
ruptured chordae, particularly of the anterior leaflet, we have quite a long experience on the same surgical team with the management of ruptured chordae by repair of the valve. In fact, around 1971 it was apparent that patients who had repair of the posterior leaflet following rupture of its chordal support were doing much better than those who had repair of the anterior leaflet. Therefore, it has been our policy since 1971 to repair only those valves that have lost chordal support to the posterior leaflet. Since 1971, 40 patients have undergone such a repair and only one has subsequently required mitral valve replacement. Finally, I would like to thank Dr. Duran and Dr. Carpentier and Dr. Kay for their helpful remarks, which will enable others to obtain optimum results from valve repair.
Copyright information The appearance of a code at the bottom of the first page of an original article in this journal indicates the copyright owner's consent that copies of the article may be made for personal or internal use, or for the personal or internal use of specific clients. This consent is given on the condition, however, that the copier pay the stated per copy fee through the Copyright Clearance Center, Inc., P.O. Box 765, Schenectady, N. Y. 12301,518-3744430, for copying beyond that permitted by Sections 107 or 108 of the U. S. Copyright Law. This consent does not extend to other kinds of copying, such as copying for general distribution, for advertising or promotional purposes, for creating new collective works, or for resale.
THORAC CARDIOVASC SURG
79:349-357, 1980
Results of mitral valvuloplasty with a suture plication technique Between January, 1975, and October, 1978, a total of 243 patients underwent repair of the mitral valve with a suture plication technique. Mitral valve replacement (MVR) was performed in 36 cases in which significant residual regurgitation was apparent on testing the valve after repair. The hospital mortality rate was 5.7%. Five patients required MVR within I month of operation. A total of 190 patients were discharged from hospital with what was considered to be a satisfactorily functioning mitral valve. Excluding patients from overseas, detailed follow-up information is available in 80 cases. Of these 80 patients, 13 have subsequently undergone MVR. Factors favorably affecting survival and durability of repair are degenerative valve disease and age below 55 years. Clinical and echocardiographic assessment indicate that this method of repair initially produces good symptomatic improvement and a normal or nearly normal pattern of left ventricular filling. The good early results are only maintained in patients with degenerative valve disease. We therefore no longer use or recommend this technique for elderly patients with rheumatic valve disease.
D. F. Shore, M.D. * (by invitation), P. Wong, M.D. (by invitation), and M. Paneth, M.D. (by invitation), London, England Sponsored by Mortimer J. Buckley, M.D., Boston, Mass.
In 1977 we 1 reported the early results of our first year's experience with mitral valve repair with the mitral plication suture (MPS). The MPS was used either to reduce the mitral anulus to an appropriate size or to correct regurgitation confined to the anterolateral or posteromedial half of the valve, which either preexisted operation or occurred following adequate relief of valve stenosis. The MPS is a double suture of 2-0 Ethiflex guarded with a Teflon pledget. It is placed at the margin of the central fibrous body and sutured circumferentially around the anulus in 2 to 3 mm steps to the posterior portion of the anulus, where it is tied over a second pledget (Fig. I). A second suture is placed at the opposite side if necessary. In cases of ruptured chordae of the posterior leaflet, the plication suture is continuous from one margin of the central fibrous body to the other, so that the leaflet repair will be supported posteriorly. From the Brompton Hospital, London, England. 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: M. Paneth, The Brompton Hospital, London, England SW3. 'Present address: The Hospital of the Albert Einstein College of Medicine, 1825 Eastchester Rd., Bronx, N. Y. 10461.
In an earlier report, we emphasized that the MPS was only part of an aggressive systematic approach to the conservation of the mitral valve. Success depended on each subunit of the mitral valve being individually scrutinized and specifically treated. Early follow-up (average 10 months) indicated that good clinical results could be obtained with a low operative mortality rate. In addition, the overall rate and pattern of left ventricular filling determined by echocardiography were normal or nearly normal in all patients studied. 1 Present series
Encouraged by these early results, we attempted repair of the mitral valve in 243 patients between January, 1975, and October, 1978. All patients were in Functional Class III or IV preoperatively. The total number of operations performed on the mitral valve during this same period was 423. The operation was performed in all cases through a median sternotomy. Cardiopulmonary bypass was established by standard aortic and bicaval cannulations. Initially, the body temperature was reduced to 32° C and the myocardial temperature was reduced to 15° C by selective aortic root perfusion. During the last 12 months of this study we routinely employed cardiople-
0022-5223/80/030349+09$00.90/0 © 1980 The C. V. Mosby Co.
349
The Journal of
350 Shore, Wong, Paneth
Thoracic and Cardiovascular Surgery
Table I. Associated procedures Procedure
No.
MR alone MR + AVR MR + tricuspid repairs MR + A VR + tricuspid valve repair MR + tricuspid valve repair MR + secundum ASD MR + primium ASD MR + CAVG MR + aortic valve repair
123 36 24* 14 I I I 2
3
Legend: MR, Mitral repair. AVR, Aortic valve replacement. ASD, Atrial
septal defect. CAVa, Coronary artery vein graft. 'Including one secundum atrial septal defect and one coronary artery vein graft.
Table II. Hospital deaths Cause
No.
Low cardiac output Preoperative systemic embolus Development of left atrial thrombus Secondary hemorrhage Empyemalsepticemia Uncertain
7 I I I I 3
Table III. Causes of late failure of repair Cause Mitral regurgitation Complete dehiscence of MPS Partial dehiscence following infection Partial dehiscence in the absence of infection Stretching of remaining chordae tendineae Progressive disease of subvalvular apparatus Shrinkage of valve leaflets Restenosis Hemolysis
No.
10 I I 3 I 2 2
the prolapsing area or by shortening of individual chordae." Adherent chordae were separated by sharp dissection down to and including papillary muscles. One or two MPSs were then inserted as described previously. The repair was tested by rendering the aortic valve incompetent and allowing the ventricle to fill. We have found that this maneuver produces a sufficient increase in left ventricular pressure to close the mitral valve and demonstrate the presence of any residual incompetence. When bypass was discontinued, simultaneous left atrial and left ventricular pressures were measured to exclude the presence of a significant transmitral gradient. Additional procedures were performed as indicated in Table I. Statistical evaluation. Three-year survival figures were calculated by the actuarial method." Ninety-five percent confidence limits were determined by the Greenwood formula. 4 Assessment of valve function after repair. We have found echocardiographic assessment of left ventricular filling to be a reliable means of evaluating mitral valve function after repair. Real-time M-mode echocardiograms were recorded, digitized, and computer analyzed." Linear traces of left ventricular dimension and the rate of change of cavity size were derived. The peak rate of dimension change (PRDC) of the transverse dimension of the left ventricle reflects the peak volume of blood flow through the mitral valve. Thus the PRDC can indicate whether the valve orifice is normal (10 to 20 cm/sec), stenotic « 10 cm/ sec), or regurgitant (>20 ern/sec). Results
3 I
Legend: MPS, Mitral plication suture.
gic arrest together with topical cooling of the myocardium and reduction of body temperature to 28° C. The left atrium was opened posterior to the interatrial groove and the mitral valve was inspected to assess its suitability for repair. Valves with rigid, calcified orifices, with large amounts of calcium in the chordae tendineae or papillary muscles, or with destruction of more than half of the tissue of the anterior leaflet were not considered suitable for repair. In all other cases repair was attempted. Abnormal commissures were split along their line of fusion to within 2 mm of the anulus. Leaflets were decalcified and valve clefts or perforations were sutured with interrupted sutures of 4-0 Ethiflex. Prolapsing leaflet margins were repaired by trapezoidal excision of
Hospital deaths. There were 14 deaths (5.7%) within I month of operation. The causes of death are described in Table II. All seven of the patients dying with low-output syndrome were considered to have a properly functioning valve. Two deaths followed emergency mitral valve replacement (MVR) in the early postoperative period. Operative failures. Of the 246 attempted repairs, 36 were judged to be unsatisfactory and the valve was immediately excised and replaced with a Bjork-Shiley prosthesis (Pyrolite disc size 29 or 31). Ninety percent of the operative failures occurred in patients with advanced rheumatic valve disease. Failure resulted either because it was impossible adequately to free the individual components of the subvalvular apparatus or because of significant residual regurgitation after completion of the repair. Early failure of repair. Careful clinical evaluation of mitral valve function was made postoperatively. In those cases in which there was any doubt that function
Volume 79
Mitral valvuloplasty
Number 3 March,1980
Central fibrous body
,
Anterior leaflet
'"\
/\
I I
100 (80) ..... (80)''............ 80
\
v» a.::
\ \
> > a.::
I I / / /
------(19) ......
---(19)
o 60
\ \
II
'- -
35 1
~
40
ltI!
Posterior leaflet
Fig. 1. Mitral plication suture inserted on both sides of the mitral valv~.
was satisfactory, echocardiographic assessment of left ventricular filling aided in the evaluation. Five patients required MVR in the early post-operative period, three owing to mitral regurgitation following partial dehiscence of the repair and two owing to inflow obstruction. The latter probably was not recognized at operation because the cardiac output was too low to produce a significant gradient. One hundred ninety patients were discharged from the hospital and were considered to have a satisfactorily functioning mitral valve. Long-term follow-up. Of the 190 patients discharged, 110 were from overseas and therefore were not available for detailed clinical and hemodynamic assessment. This study therefore concerns the followup of the remaining 80 patients. The average age of these 80 patients at operation was 54 years, contrasting sharply with the average age of 33 years of those patients from overseas who were not included. Of the 80 patients, 13 have subsequently required MVR. The causes of failure are listed in Table III. Seven patients in this series required MVR because of progression of rheumatic valvular disease. The patient who required reoperation for hemolysis had a very small regurgitant jet. This was readily corrected without the need for valve replacement and there was no evidence of hemolysis postoperatively. There have been nine late deaths, four following MVR, one following a cerebrovascular accident, and four of unknown cause. The overall survival rate, calculated by the actuarial method, is 89% at 3 years (Fig. 2). The chance of surviving without requiring reoperation is 72% at 3 years (Fig. 2).
20
- - % Survival
--- % Survival with Natural Valve
Fig. 2. Survival 3 years following mitral repair.
100
--
.....
80
, .....
I
',----- I
Vl
a.::
0 60
> > a.::
::::>
Vl
ltI!
±2SD 40 20
Age
-<50 ---)50
2 YEARS Fig. 3. Influence of age on survival following mitral repair.
The influences of age (Fig. 3), surgical pathology (Table IV, Fig. 4), and associated valve disease (Fig. 5) on postoperative survival were evaluated statistically by comparing survival curves after the entire group of patients had been divided into appropriate groups. At 3 years, significantly adverse correlates of survival were age above 50 years and rheumatic valvular stenosis. Associated valve disease did not adversely affect survival. Patients with degenerative valve disease had a significantly improved chance of event-free survival when compared with patients with rheumatic valvular disease, either predominantly stenotic or predominantly regurgitant (Fig. 6). Incidence of thromboembolism. Patients were not given anticoagulants unless aortic or tricuspid valve
The Journal of Thoracic andCardiovascular Surgery
352 Shore, Wong, Paneth
,,
.~------i-I "
80 V)
a.:
0
> > a.:
::::>
V)
60
",----1
±2 SD 40
~
20
w w 80 a.:
LL.
f--
zw
60 > w V)
a.:
Aetiology --- Rheumatic Incompetence _. _. Rheumatic Stenosis
2
.....
,,
,
......
-----_...:..~
,,
. !. I
"
±2SD
Aetiology ~ 40 > --- Rheumatic Incompetence a.: ::::>
V)
~
- - Degenerative
,,
20
_. _. Rheumatic Stenosis - - Degenerative
3
YEARS
Fig. 4. Influence of surgical pathology on survival following mitral repair.
80
-.. . . . . . . . . . -----.:1
V)
a.:
o
> > a.:
::::>
V)
60 ±2SD 40
- - - Multiple Valve Disease --Mitral Valve Disease Only
20
Fig. 6. Influence of surgical pathology on survival without the need for subsequent valve replacement following mitral repair. Table IV. Etiology
Rheumatic Degenerative Infection (rheumatic valve) Infection (normal valve) Coronary artery disease Congenital Uncertain
MS
MR
19
25 19 5 I
1 I 8
Legend: MS, Mitral stenosis. MR. Mitral repair.
Table V. Echocardiography (49 patients) PRDC Normal range (10-20 em/sec) Stenotic «10 em/sec) Regurgitant (>20 em/sec)
I No. of patients 34
12 3
Fig. 5. Influence of associated valvular disease on survival following mitral repair.
Legend: PRDC. Peak rate of dimension change.
replacement had been required. Definite evidence of thromboembolic disorders is available in two cases. One patient died following a cerebral embolus and a second patient developed a right hemiplegia and has residual weakness. Clinical and hemodynamic assessment. Of the 71 patients still alive, 62 have not required valve replacement. All but five of these patients were symptomatically improved following operation. Eight patients improved by three grades, 34 patients by two grades, and 15 patients by one grade. The condition of 49 patients, including four of the
five patients who were not symptomatically improved was assessed by echocardiography (Table V). Thirtyfour patients had a peak rate of dimension change (PRDC) within the normal range (Table V). Twelve patients had a PRDC below 10 em/sec, but only two of these patients had a PRDC below 7 em/sec. All the patients with a PRDC below 10 em/sec were in Functional Class I or II. The three patients who had a PRDC in the regurgitant range were in Functional Class II. Of the five patients who were not symptomatically improved following operation, all had a PRDC within normal range.
Volume 79 Number 3 March. 1980
Eleven of the 15 patients with a PRDC outside the normal range had rheumatic valve disease; four of these patients had predominantly stenotic lesions and seven, predominantly regurgitant valves. Seventeen patients with degenerative valve disease were examined by echocardiography. Only one of these patients had an abnormal PRDC of 23 ern/sec, indicating slight mitral regurgitation. Discussion
A durable and easily reproducible method of conserving the diseased mitral valve remains an attractive goal. The advantages of repair over prosthetic replacement are a low incidence of thromboembolic episodes without, in most cases, the need for anticoagulation, and the retention of a valve whose function is normal or more nearly normal than the best of available prostheses. However, to be acceptable, any method of repair must have an operative mortality rate and postoperative event-free survival rate which compare favorably with those following valve replacement with either mechanical prostheses'":" or tissue valves.?: 10 The operative mortality rate (5.9%) and 3 year survival rate (89%) in our series indicate that this method of repair satisfies these criteria. Repair with the MPS promised to be quick and simple to perform. If unsuccessful, the valve could then be readily excised and replaced without much loss of time. These factors, together with the fact that the valve ring retained the normal physiological ability to shorten its circumference during ventricular systole," led us to believe that the MPS technique would prove an advantage over methods of repair which depended on the use of annuloplasty rings. The high incidence of operative failures was due in part to the broad criteria for selection, which included patients with advanced rheumatic disease provided only that extensive calcification was absent. In addition, the technique we used for testing the valve may have overestimated the number of patients with significant residual regurgitation. In some of these cases the regurgitation may have been less severe had competence of the valve been assessed with a contracting ventricle, physiological shortening of the mitral ring circumference, and normal development of intraventricular pressure. However, one of the advantages of such overestimation of the number of failed repairs was that there were no hospital deaths in this group of 36 patients. A method of testing repair procedures that comes close to the ideal has been described by Nair and assoelates." An appraisal of the early results of this technique
Mitral valvuloplasty
353
suggested that predictably good symptomatic results could be obtained together with a normal or nearly normal rate and pattern of left ventricular filling. This is in contrast to the situation after MVR with a BjorkShiley or Starr-Edwards prosthesis, which produced both an abnormal rate (PRDC; Bjork-Shiley = 7 to 8 em/sec; Starr-Edwards = 6 ern/sec) and an abnormal pattern of left ventricular filling. However, the number of patients who have required MVR within 3 years of the initial operation is disappointing. We have been unable to match the durability of the repair that has been obtained with the use of annuloplasty rings. In addition, Duran and Ubago" compared left ventricular function following the use of a rigid ring and a flexible ring and found no significant difference. Therefore, the theoretical advantage of retaining ring flexibility may in practice be minimal. Although dehiscence was responsible for failure of the repair in four patients in the absence of infection, progression of the disease was responsible in a further eight patients. Seven of these eight patients had rheumatic valve disease. Progression of the disease is rarely cited as a cause of failure in other series. However, one difference, which may have had a significant influence in this respect, is the average age of 54 years of patients in our series compared with the average age of patients in the series reported by Duran and Ubago.P 37 years, and Carpentier and associates ,14. 15 40 and 43 years. As pointed out earlier, those patients from overseas whose average age was 33 years could not be included in the long-term follow-up. However, as a result of the high incidence of MVR following repair, we have abandoned attempts to conserve the mitral valve in the elderly patient with rheumatic valve disease except in those cases in which only a commissurotomy is required. On the other hand, repair of the degenerative valve has been followed by a 94% survival at 3 years and a low incidence of subsequent MVR. All 19 of the patients in this study with degenerative valve disease had ruptured chordae; 18 patients had rupture of chordae to the posterior leaflet and one patient, rupture of chordae to the anterior leaflet. In addition, there was redundancy of that part of the affected leaflet which had lost its chordal suppost. Slight-tosevere mitral ring dilatation was also present. It is our practice in these cases to perform a trapezoidal excision of the redundant leaflet tissue. After excision, the cut edges of the leaflet are approximated by interrupted sutures of 4-0 Ethiflex. The MPS is inserted to narrow the ring and support the leaflet repair. In our experience progression of the disease is rare in this group; stretching of the remaining chordae was
The Journal of Thoracic and Cardiovascular Surgery
354 Discussion
responsible for recurrent regurgitation necessitating MVR in only one case. The symptomatic results and the hemodynamic status of this group of patients is excellent following operation. The results are clearly superior to those obtained following repair of rheumatic valves, because leaflet pliability and the function of the subvalvular apparatus are normal following repair. Conclusion
The operative mortality rate (5.9%) and 3 year survival rate (89%) after repair of the mitral valve with a suture plication technique compare favorably with those after MVR. The restoration of normal mitral valve function after repair, demonstrated by echocardiography in the early postoperative period, is not maintained in patients with rheumatic valvular disease. Deterioration of function is more common in patients over 55 years of age. This has resulted in 16% of patients requiring MVR within 3 years. We therefore no longer use or recommend the use of this technique for repair of the valve in the elderly patient with rheumatic valve disease. Good functional results are maintained in patients with degenerative valve disease. The use of this technique in younger patients with rheumatic disease requires further evaluation.
2
3 4
5
6
7
8
REFERENCES Burr HB, Krayenbuhl C, Sutton MStJ, Paneth M: The mitral plication suture. J THORAC CARDIOVASC SURG 73:589-595, 1977 McGoon DC: Repair of mitral insufficiency due to ruptured chordae tendineae. J THoRAc SURG 39:357-362, 1960 Grunkemeier GL, Starr A: Actuarial analysis of surgical results. Ann Thorac Surg 24:404-408, 1977 Cutler SJ, Ederer F: Maximum utilization of the life table method in analyzing survival. J Chron Dis 8:669-672, 1958 Gibson D, Brown DJ: Measurement of instantaneous left ventricular dimension and filling rate in man using echocardiography. Br Heart J 35:1141-1149,1973 Starr A: Mitral valve replacement with ball valve prosthesis: A current appraisal of late results, The Mitral Valve, D Kalmanson, ed., Acton, Mass., 1976, Publishing Science, Inc., Publisher, p 413-429 Bjork VO: Results of mitral valve replacement with the Bjork-Shiley tilting disc valves, The Mitral Valve, D Kalmanson, ed., Acton, Mass., 1976, Publishing Science, Inc., Publisher, p 449-455 Sutton MStJ, Miller GAH, OIdershaw P, Paneth M: Valve replacement with the Bjork-Shiley valve. Experience with 390 patients. Read at the Summer Meeting, The
Thoracic Society, Liverpool, England, July 10, 1976 9 Ionescu MI, Tandon AP, Mary DAS, Abid A: Heart valve replacement with the Ionescu-Shiley pericardial xenograft. J THoRAc CARDIOVASC SURG 73:31-42, 1977 10 Angell WW: Long term results of tissue valve grafts for mitral replacement, The Mitral Valve, D Kalmanson, ed., Acton, Mass., 1976, Publishing Science, Inc., Publisher, p 481-488 II Tsakiris MD: The physiology of the mitral annulus, The Mitral Valve, D Kalmanson, ed., Acton, Mass., 1976, Publishing Science, Inc., Publisher, p 21-26 12 Nair KK, Yates AK: Direct evaluation of mitral valve function during surgery following conservative procedures. J THORAC CARDIOVASC SURG 73:684-685, 1977 13 Duran CG, Ubago JLM: Clinical and hemodynamic performance of a totally flexible ring for atrioventricular valve reconstruction. Ann Thorac Surg 22:458-463,1976 14 Carpentier A, Reiland J, Deloche A, Fabiani J, D'AIlaines C, Bordeau P, Piwnica H, Chauvaud S, Dubost C: Conservative management of the prolapsed mitral valve. Ann Thorac Surg 26:294-302, 1978 15 Carpentier A: Plastic and reconstructive mitral valve surgery, The Mitral Valve, D Kalmanson, ed., Acton, Mass., 1976, Publishing Science, Inc., Publisher, p 527-540
Discussion (Papers by Reed [page 321), Duran [page 326], Carpentier [page 338], Shore [page 349], and their colleagues)
DR. ROBERT W. M. FRATER Bronxville. N. Y.
It appears to me that there are three problems with plastic repairs of the mitral valve. The first is simple: "Life is short and the art is long" (Hippocrates). The surgeons who have talked today are clearly artists, and Dr. Carpentier, the doyen of this school of artists. They have spent the time and devoted the intense effort that is necessary for the development of artistic skill. Not all of us have done that, and therefore not all of us can remotely hope to achieve their results. The second problem is this: How can one be sure that the valve is neither stenotic nor insufficient at the end of the repair before closing the atrium? Drs. Reed and Carpentier particularly, and the others, too, have suggested that this may be possible using their artistic skill. Third, do the repairs really hold up with time? The evidence from the papers herein presented from four different countries is that some of the repairs certainly do seem to hold up with time. We do have the problem of comparing apples with oranges-young rheumatic patients with pure insufficiency, older rheumatic patients with mixed stenosis and insufficiency, and older patients with degenerative valvular disease. From what we have learned today, it is very clear that the young rheumatic patient with pure insufficiency does well. However, Dr. Shore's paper suggests very strongly that
Volume 79
Discussion
Number 3 March,1980
valvuloplasty in the elderly rheumatic patient, particularly the one with insufficiency and stenosis, does not hold up well, whereas valvuloplasty in the elderly patient with degenerative disease does. Regarding the second question-how to be sure that the right results have been achieved at the time of operation-I would like to suggest that one of the most difficult problems facing the surgeon at the operating table is to decide how to shorten elongated chordae so as to produce a competent valve. Dr. Carpentier has described a technique for this which he clearly executes very well. However, the art lies in deciding exactly how much the chordae should be shortened. There is a way of deciding this provided that only one cusp has elongated chordae. The rules are very simple: Both cusps hang to the same level in the ventricle, and the distance from a point on a papillary muscle to the true edge of either cusp is always the same and always less that the distance to the atrioventricular ring. Given two cusps, one with elongated chordae and one with chordae of normal length, if the free edge of the prolapsing cusp is stitched temporarily to the free edge of the nonprolapsing cusp and the cusps are gently pulled on to put the shorter group of chordae under slight tension, it then seems possible to perform Dr. Carpentier's maneuver knowing full well that the end result will be correct.
DR. JEROME HAROLD KAY Los Angeles, Calif.
I am so pleased that other surgeons are reporting good results with mitral repair for mitral insufficiency, since I have been advocating this for the past 20 years. The incidence of peripheral emboli is far lower with repair than with mitral valve replacement with the porcine valve. One hundred sixty-five patients had mitral repair for pure mitral insufficiency from I to 20 years ago. Fifteen patients were reoperated upon for recurrent mitral insufficiency or mitral stenosis. I am quite sure this figure is far lower than the number of patients requiring reoperation following mitral replacement with a porcine valve. Regarding the surgical technique, we have always used sutures alone, never mitral rings. I feel strongly that the use of a ring is unnecessary for the surgical treatment of pure mitral insufficiency. All patients with pure mitral insufficiency have a dilated anulus , and 60% of the patients also have tom chordae tendineae or ruptured papillary muscle. For repair, the leading edge of the valve at the area of the tom chordae tendineae is sutured down to the papillary muscle nearest to this area. If an anterior leaflet has the tom chordae tendineae, that part of the valve has to be sutured to the papillary muscle. If the mural leaflet has the tom chordae tendineae, the repair has to be performed so as to allow the anterior leaflet to balloon out and abut against the mural leaflet. This is extremely important in the technique. With tom chordae tendineae of the mural leaflet, the area of the leaflet with the tom chordae tendineae is sutured to the papillary muscle nearest the tear.
355
Seventeen percent of the patients have stretched chordae tendineae along with the dilated anulus, and 23% have a dilated anulus only. For the repair, the area of the mural leaflet with the stretched chordae tendineae should be sutured down to the nearest papillary muscle. One may have to suture the mural leaflet down to the papillary muscle in two or three different areas. If the anterior leaflet has stretched chordae tendineae, the anterolateral part of the anterior leaflet has to be sutured to the anterolateral papillary muscle to allow the anterior leaflet to balloon out. In all of the patients with pure mitral insufficiency, whether or not there are stretched chordae tendineae, there is always a dilated anulus. In repairing the dilated anulus, sutures are placed at the posteromedial and anterolateral commissural areas to do away with approximately 70% of the mural anulus. Care must be taken not to encroach on the anulus of the anterior leaflet. Replacement is far easier than repair in most patients; however, with experience, repair is feasible in approximately 75% of the patients and is far better for the patient than replacement. MR. MARK BRAIMBRIDGE Landon, England
As I see it, now that most of us are using cardioplegia, the problem in mitral valve repair is testing the valve. I would like to ask each of the discussers to tell us how they test their valves and what constitute their immediate criteria of successful repair. In our experience, when the ventricle is completely relaxed it is difficult to know how the tone in the valve ring will be when the heart is once again reperfused. We dilate the ventricle gently with cold saline through the left ventricular vent to assess the competence of the valve. This method is very empiric, but it does seem to give us a fair idea of how the valve will be after the operation. I would be most interested to know how the participants test the valves. DR. JOHN W. KIRKLIN Birmingham. Ala.
A question directed toward Dr. Carpentier has come to the desk from the audience. It is as follows: At reoperation upon a valve that has been repaired with a ring, is the valve now so small because of the ring that it is difficult to insert a prosthesis of adequate size? I would like to ask a question of each of the essayists: Is it a particular problem if a patient with mitral incompetence has ruptured chordae or very elongated chordae to the anterior leaflet? DR. R E ED (Closing) I was pleased to see that despite the variety of techniques used, the results were uniform. The operative mortality rate for this group may be explained partially by their youth and by the selectivity in excluding patients with heavy calcification. The very low incidence of thromboembolism is similar
The Journal of
35 6 Discussion
Thoracic and Cardiovascular Surgery
irrespective of technique and is considerably lower than that reported for valve replacement. It was good to learn of Dr. Kay's results in patients with coronary artery disease. I have followed his work with great interest and respect over the years and often have felt that he was the only other voice in the wilderness. I want to thank Dr. Frater for his kind remark, although I can take very little credit for being an artist. We have tried to describe a technique that has mathematical precision and is readily duplicated. It can be learned easily, but meticulous attention to detail is necessary to make it work. To answer Mr. Braimbridge's question: We deliberately make the valve smaller than is probably necessary but, by so doing, almost ensure that it will be competent. The valve is tested routinely by coming off bypass and elevating the pressure with a systemic pressor. This increase in peripheral resistance is necessary to assure that the valve is being appropriately stressed. During this time we palpate for regurgitation through an untied suture in the left atrial closure. Yes, Dr. Kirklin, the most troublesome patients are those with elongated chordae to a prolapsed anterior leaflet. This condition is treated by excising a wedge of tissue and then bringing together valve edges that are adequately supported by chordae which are not stretched. Most of this free edge no longer participates in the line of closure because the annuloplasty repair advances the major leaflet to form a new line of closure. All patients have annuloplasty regardless of whether they have had a valvuloplasty procedure. DR. 0 U RAN (Closing) Regarding the technical difficulties of performing this type of conservative operation, I would like you to think for a moment how difficult we would consider these procedures if mitral replacement had been developed before mitral commissurotomy. How many commissurotomies would be done today? Although in our unit the residents learn first to replace a valve and then to conserve it, it is not a matter of artistry but of craftmanship. A chronologie study of the percentage of valve replacements versus conservation, per individual surgeon of our unit, shows that it takes some time for the newly arrived to this technique to be convinced of its usefulness. A steady rise in the number of annuloplasties is seen with time. I would therefore be glad if you will just evaluate the next mitral valve before taking it out. Our method of checking valve competence intraoperatively consists of filling the left ventricle directly through the left vent, which is temporarily connected to the arterial line. It was published in The Annals a/Thoracic Surgery. One of our surgeons thought that, under cardioplegic conditions, we might be getting too high a rate of insufficiency. Back in the animal laboratory, we saw that the normal mitral valve in a flaccid heart did not regurgitate. Finally, I will address the question of more complex techniques, such as chordal shortening. We have performed only
18 such operations and, although very satisfied with this technique, I think the experience is too short for a comment. Such operations are difficult and take time. We therefore keep it simple with a straightforward annuloplasty after a good commissurotomy.
DR. CAR PEN TIE R (Closing) There have been many interesting questions. In answering Dr. Frater's question, we have demonstrated that the repair holds up with time. His suggestion of suturing the free edges of the leaflet to assess the degree of shortening is an interesting idea, and I congratulate him. Dr. Kay has been a pioneer in this field, and it is appropriate for him to remind us that elongated chordae occur more frequently than expected. However, I do not share his opinion that some sort of anulus remodelling is not necessary when the anulus is dilated. Mr. Braimbridge raised the question of the difficulty in assessing valve function when using cardioplegia. This is the reason why cardioplegia should be instituted only after valvular function has been evaluated and the various lesions have been analyzed carefully. Dr. Kirklin touched on the problem of the prolapsed anterior leaflet. If the cause of the prolapse is a chordal rupture, only rupture of one or two chordae can be treated by leaflet resection. Rupture of more than two chordae would require extensive resection of the anterior leaflet and therefore a lack of tissue. If the prolapsed leaflet is due to elongated chordae, the condition can be treated by chordal shortening. The degree of shortening is determined by measuring the degree of overriding of the free edge of the anterior leaflet above the plane of the orifice. Finally I would like to elaborate on two limiting factors which may explain why surgeons react differently when faced with the same lesion. The first is difficulty in patient selection. Patient selection depends on valvular lesions. Valvular lesions can be diagnosed prior to the operation, but most accurately during operation. It is critical for the surgeon to be able to recognize all of the lesions. Since there is only one technique for each lesion. the problem is not to decide which technique should be used but rather to recognize the lesion. The second point is reproducibility of the techniques. This depends upon the will of the surgeon to train himself in order to acquire the necessary experience. It also depends upon our ability to simplify the techniques. In this regard, and answering Dr. Duran's question, we have recently been able to simplify chordal shortening. The details of the modified technique are presented in the paper. Reproducibility of the techniques also depends on the quality of exposure. Good exposure is obtained by (1) lysis of any adhesions present, (2) extensive incision of the atrium, (3) placing a pad in the pericardium to expose the papillary muscles, and (4) using a self-retaining retractor.
Volume 79
Discussion
Number 3 March,1980
DR. S H 0 R E (Closing)
To answer Mr. Braimbridge's question, we tested the mitral valve after repair by rendering the aortic valve incompetent and allowing the ventricle to fill. We also took care to measure left ventricular and left atrial pressures simultaneously at the discontinuation of cardiopulmonary bypass to exclude the presence of a significant transmitral gradient. Five patients required valve replacement in the early postoperative period in the first year of our experience with this technique, but since then only one patient has required valve replacement in the early postoperative period. These results indicate that we are getting better at determing whether or not the valve repair has been adequate. To answer Dr. Kirklin's question about the management of
35 7
ruptured chordae, particularly of the anterior leaflet, we have quite a long experience on the same surgical team with the management of ruptured chordae by repair of the valve. In fact, around 1971 it was apparent that patients who had repair of the posterior leaflet following rupture of its chordal support were doing much better than those who had repair of the anterior leaflet. Therefore, it has been our policy since 1971 to repair only those valves that have lost chordal support to the posterior leaflet. Since 1971, 40 patients have undergone such a repair and only one has subsequently required mitral valve replacement. Finally, I would like to thank Dr. Duran and Dr. Carpentier and Dr. Kay for their helpful remarks, which will enable others to obtain optimum results from valve repair.
Copyright information The appearance of a code at the bottom of the first page of an original article in this journal indicates the copyright owner's consent that copies of the article may be made for personal or internal use, or for the personal or internal use of specific clients. This consent is given on the condition, however, that the copier pay the stated per copy fee through the Copyright Clearance Center, Inc., P.O. Box 765, Schenectady, N. Y. 12301,518-3744430, for copying beyond that permitted by Sections 107 or 108 of the U. S. Copyright Law. This consent does not extend to other kinds of copying, such as copying for general distribution, for advertising or promotional purposes, for creating new collective works, or for resale.