(Circulation. 1995;92:150-154.)
© 1995 American Heart Association, Inc.
Articles |
From the Section of Cardiovascular Surgery, Mayo Clinic and Mayo Foundation, Rochester, Minn.
Correspondence to John A. Odell, MB, ChB, Division of Thoracic and Cardiovascular Surgery, Mayo Clinic, 200 First St SW, Rochester, MN 55905.
| Abstract |
|---|
|
|
|---|
Methods and Results A consecutive series of all patients (n=418; median age, 67 years) having mitral valve repair performed by two surgeons was analyzed to determine the outcome of three different annuloplasty techniques: commissural annuloplasty (n=124), complete ring annuloplasty (n=113), and an unmeasured, posterior, partial ring annuloplasty (n=181). Intraoperatively, before repair, severity of mitral regurgitation as measured by double sampling dye curves and transesophageal echocardiography was similar in all three groups; after mitral valve repair, intraoperative assessment showed a similar degree of reduction in regurgitation in the three annuloplasty groups. Before hospital dismissal, transthoracic echocardiography demonstrated that the mean mitral valve areas and gradients were similar in the three groups; more patients having commissural annuloplasty were classified as having grade II or greater regurgitation. Mortality (n=7, 1.7%) and need for reoperation (n=8, 1.9%) was low in all groups despite the fact that additional procedures were performed in 48.8% of patients. Durations of cardiopulmonary bypass and aortic cross-clamping were significantly less in patients having commissural or posterior annuloplasties compared with those receiving a complete ring annuloplasty.
Conclusions These early results indicate that the posterior annuloplasty method is reproducible and expeditious. Postoperative valve function as assessed by degree of regurgitation, transvalvular gradient, and valve area was similar to that obtained by measured, complete ring annuloplasty and superior to that found in patients having commissural annuloplasty.
Key Words: mitral valve surgery valvuloplasty
| Introduction |
|---|
|
|
|---|
The purpose of this study was to analyze the early and intermediate-term results of three different annuloplasty methods used during mitral valve repair in a recent series of 418 consecutive patients. Outcome measures included valve function (determined intraoperatively by double sampling dye curves and transesophageal echocardiography and postoperatively by transthoracic echocardiography) as well as patient survival and need for reoperation.
| Methods |
|---|
|
|
|---|
Valve Assessment
Mitral valve function was assessed by
echocardiography using Doppler and color flow
mapping to quantify regurgitation. The degree of
regurgitation was quantified on a scale of 0 to 4 as
described by Sheikh et al.13
Transesophageal echocardiography
and double sampling dye curves14 were used
intraoperatively to quantify valvular
regurgitation. Quantification by double sampling dye
curves involves injection of indiocyanine green into the left ventricle
with simultaneous and continuous withdrawal of blood from
needles placed in the left atrium and ascending aorta. Concentration of
dye determined by a Gilson densitometer is plotted against time, and
the area under the curves is computed and a regurgitant index
calculated. Studies are performed before and after repair.
Patient Assessment
Follow-up data were obtained by direct
evaluation of the
patient, telephone contact, and/or letters from patients or their
physicians; this information along with preoperative and intraoperative
variables were entered into a standard database.
Statistical Methods
Patient characteristics are summarized
with percentages for
categorical data and with means for continuous (ordinal) data. Group
comparisons were analyzed by
2 test or
ANOVA with Bonferroni t tests for multiple groups. A
probability value less than .05 was considered significant.
| Results |
|---|
|
|
|---|
Twenty-one patients had previously undergone cardiac surgery. Fifteen patients had coronary artery bypass grafting (CABG), one had CABG plus aortic valve replacement (AVR), one had atrial septal defect closure with mitral valve repair, one had atrial septal defect closure, and three patients had previous mitral valve repair. Two of these patients had two previous cardiac procedures.
Thirty-two patients had previously treated endocarditis; none were managed by repair in the acute phase. Patients older than 45 years had preoperative evaluation of their coronary arteries, and significant obstructive lesions were bypassed at the time of mitral valve repair.
Operative Findings
Intraoperative findings are summarized in
Table 1
.
The most common cause of mitral regurgitation was
ruptured chordae tendineae (n=190); in many patients, this was found in
association with elongation of other chordae and/or myxomatous disease
in the leaflets. Other common findings were prolapse of one or both
leaflets (n=119) and annular dilation with no leaflet abnormalities
(n=44).
|
Methods of Valve Repair
Unsupported segments of the posterior
leaflets were treated by
resection or plication (McGoon technique).15 Anterior
leaflet prolapse was managed by plication in 97 patients; resection of
the anterior leaflet generally was avoided. Rarely was shortening or
reattachment of anterior leaflet chordae required (15.9% in the ring
group and 7.7% in the posterior annuloplasty group).
The approach to annuloplasty evolved over time. In the earliest part of this experience, commissural narrowing was favored; later, a measured annuloplasty with complete ring was used. More recently, we have preferred an unmeasured posterior annuloplasty using two thirds of a 27-mm flexible Duran ring. In selected patients, single or bicommissural narrowing is still used.
There were 124 patients who had
either a unicommissural or
bicommissural annuloplasty; 113 patients had an annuloplasty using a
measured complete ring (Carpentier-Edwards, n=78; Duran, n=35;
mean
size, 31.7 mm±2.7 mm [SD]), and a posterior annuloplasty was
used in
181 patients. Details of valve repair are summarized in Table
2
.
|
Assessment of Valve Function and Repair
Intraoperative
transesophageal
echocardiography was performed in 255 patients (59
commissural, 35 ring, and 161 posterior annuloplasty). Before valve
repair, the percentage of patients having grade III or IV
regurgitation was 85.2% in the commissural group,
91.4% in the ring group, and 97.5% in the posterior annuloplasty
group (P<.001 versus commissural). After valve repair,
transesophageal echocardiography
demonstrated grade III or IV regurgitation in 1.8% of
patients with commissural annuloplasty, 2.5% of patients with measured
ring annuloplasty, and 1.2% of patients with posterior annuloplasty.
The percentage of patients with grade 0 or grade I
regurgitation was 81.8%, 84.9%, and 72.7%,
respectively, for the three groups, and these differences were not
significantly different (Table 3
).
|
Transthoracic echocardiography (TTE) was performed before hospital dismissal in 379 patients. The percentage of patients with residual grade III or IV regurgitation was 8.6%, 3.8%, and 4.5% in the commissural, ring, and posterior annuloplasty groups, respectively, and the differences were not significant. The percentage of patients with mitral regurgitation grade 0 or I by TTE was 60.2%, 84.9%, and 72.7% (P<.0001 commissural versus complete or partial ring).
The mean regurgitant index measured before repair was 47.7% for the commissural group (n=77), 52.6% for the ring group (n=93), and 54.9% for the posterior annuloplasty group (n=139). Postrepair regurgitant indices were 2.4%, 1.5%, and 1.2%, and there was no statistical difference between the groups either before or after surgery.
Operative Variables
The average (±standard deviation)
durations of
cardiopulmonary bypass and aortic cross-clamp were
82.5±38.7 and 49.1±22.9 minutes for the commissural group,
100.5±35
and 67.8±23.1 minutes for those having a complete ring, and
71.4±34.6
and 49.5±21 minutes for the posterior annuloplasty group
(P<.001 for complete ring versus commissural or posterior
annuloplasty groups). Additional cardiac procedures may account for
some of the observed differences. Analysis of patients who had
mitral repair with no associated procedures demonstrated no difference
in duration of bypass or cross-clamp comparing the commissural and
posterior annuloplasty groups; however, bypass and cross-clamp
times were significantly longer for patients having measured complete
ring annuloplasties.
The additional surgical procedures performed in
each group are listed
in Table 4
. Seventy-seven additional cardiac
procedures were done in the commissural group (62%, P<.001
versus posterior annuloplasty), 52 in the ring group (46%), and 75 in
the posterior annuloplasty group (41.4%). Additional noncardiac
procedures were completed in eight patients; these included thymectomy,
repair of pectus excavatum deformity, pulmonary lobectomy, and
mastectomy. Four patients had concomitant lung biopsy at the time of
mitral valve repair.
|
Mortality and Reoperation
Seven patients (1.7%) died within
30 days or within the same
hospitalization: 5 in the commissural group, none in the complete ring
group, and 2 in the posterior annuloplasty ring group. All except one
of the patients who died had an additional procedure (CABG in 5, AVR in
1). Death was due to pulmonary embolism in two patients; other
causes of early mortality were arrhythmia on the third
postoperative day, failure to wean from bypass in a patient with poor
ventricular function preoperatively, aortic dissection
after intra-aortic balloon placement, stroke on the third
postoperative day, and myocardial infarction. One patient who had a
commissural repair and posterior leaflet resection required reoperation
and mitral valve replacement four days after repair for dehiscence of
the repair. She died suddenly 8 days later and had postmortem evidence
of coronary and pulmonary embolism.
Ninety-six percent of patients were followed after discharge; the mean follow-up in the overall patient population was 2.1±2.1 years (SD), with a range of 0 days to 10.1 years. In the commissural group, the mean follow-up was 2.9±2.3 years (range, 0 days to 10.1 years); in the ring group, 3.5±2.0 years (range, 0 days to 9.2 years); and in the posterior annuloplasty group, 0.7±0.6 years (range, 0 days to 3.8 years). Eight patients required late reoperation and mitral valve replacement; one patient (0.8%) had commissural annuloplasty, 4 (3.5%) had measured, complete ring annuloplasty, and 3 (0.6%) had posterior annuloplasty. The linearized rates of reoperation for the commissural and complete rings, respectively, were 0.29 and 1.03 per 100 patient-years. The follow-up in the posterior annuloplasty group is insufficient for a meaningful linearized rate. The reasons for reoperation were persistent hemolysis in one patient having prior commissural repair, bacterial endocarditis caused by Haemophilus sp in one patient in the posterior annuloplasty group and progression of the underlying mitral valve disease in three patients (two in the ring group and one in the commissural group). One patient had mitral valve replacement elsewhere, and details of the operative findings are unknown.
Reoperation was necessary for technical reasons in two patients having complete ring annuloplasty. In one patient, sutures had torn through the annulus of the anterior leaflet; in the other, the previous plication of the anterior leaflet had loosened.
| Discussion |
|---|
|
|
|---|
The remainder of the mitral annulus thins as it separates from the fibrous trigones, and because of the attachment to the base of the left ventricle, both the size and shape change during the cardiac cycle; it is the posterior portion of the annulus that lengthens when the annulus dilates. Thus, mitral valve annuloplasty should reduce the length of the posterior annulus and not distort the intertrigone or straight portion of the mitral annulus to which the anterior leaflet is attached.
Previously, Wooler et al,10 Kay et al,16 and Reed et al17 described methods to reduce the posterior annulus by plicating the annulus at the commissures; small, mainly anchoring sutures were placed on the anterior side of the annulus at the commissure with considerably larger bites on the posterior annulus, which effectively decreased the length of the posterior annulus.
In 1971, Carpentier et al11 described a method of narrowing the annulus with a rigid ring, which more evenly distributes tension along the posterior annulus. Duran and Ubago12 subsequently introduced a flexible complete ring, which follows a similar principle and has the theoretical advantage of allowing normal flexing of the annulus during the cardiac cycle. Both types of complete rings are sized by measurement of the intertrigone distance and/or the area of the anterior leaflet of the mitral valve. A recent development that allows adjustment of the annulus after the sutures have been placed around the complete annulus has been described.18
The impetus to use the posterior annuloplasty arose from two sources. First was the understanding that the anterior annulus plays no part in the dilation of the mitral apparatus, and by eliminating that part of the repair, the procedure was simplified and shortened. Second, there were early reports that a complete ring may distort the anterior leaflet and contribute to outflow tract obstruction and systolic anterior motion of the mitral valve.19 This earlier belief is no longer accepted. Additionally, the insertion of sutures along the anterior annulus is potentially injurious to the anterior leaflet and the underlying aortic valve.
Some surgeons have created posterior annuloplasties with continuous suture (deVega type),20 21 22 23 a portion of 4-mm polytetrafluoroethylene (PTFE),24 or autogenous pericardium.20 25 26 We follow this concept, but unlike others, we have tried to make repair of the mitral valve as simple as possible. No attempt is made to determine the optimum size to which the posterior annulus should be narrowed; instead, a size 27 Duran flexible ring is simply divided to two thirds of its circumference (5.6 cm). Cooley et al27 also use a standard size ring fashioned from a portion of Dacron graft for mitral valve repair.
Continued use of different annuloplasty techniques suggests that the various methods yield roughly equivalent results. In fact, mitral valve regurgitation was reduced to a similar degree in each of the annuloplasty groups in this study. However, some early deterioration in valve function was evident in patients receiving commissural annuloplasty, in which 81.8% of patients had grade I or less regurgitation by intraoperative TEE compared with 60.2% having this grade of mitral regurgitation at the time of hospital dismissal.
Although the mitral valve of some patients can be managed without annular reinforcement,28 we believe that some form of annuloplasty is necessary in every patient having mitral valve repair.29 In addition to correcting existing annular dilation, a properly constructed annuloplasty prevents further dilation, normalizes leaflet coaptation, and reinforces leaflet suture lines. Comparisons have been made of rigid and flexible annuloplasty rings2 and ring and commissural annuloplasty,22 but to our knowledge, no systematic evaluation of the three types has been made from the same institution.
Two clear limitations of our study should be recognized. First, the patients were not randomized; this probably accounts for the finding that patients having commissural annuloplasty had slightly less severe grades of regurgitation before surgery. We continue to use this method of annuloplasty for selected patients who have mild or moderate degrees of mitral regurgitation, no leaflet abnormalities, and mild dilation of the annulus. Also, it can be argued that some of the group differences in intraoperative variables might be explained by improvement in the learning curve.
A second limitation is that duration of follow-up is relatively short, particularly for patients with the posterior annuloplasty. We believe, however, that technical imperfections would be manifest on echocardiography early after operation; in clinical follow-up there has been no difference in need for reoperation when the three groups were compared.
Despite these potential shortcomings, our review indicates that the unmeasured posterior annuloplasty is effective, reproducible, and expeditious, requiring little more time than simple commissural plication. The posterior annuloplasty obviates the need to fully expose the intertrigone area for measurement of ring size, and there is no need to place sutures in the anterior annulus, a maneuver that can distort the valve.
Advocates of an adjustable annuloplasty ring describe the need to "fine tune" the size of the annulus in 9 of 21 patients,18 and they also stated that the two popular and currently available annuloplasty rings require accurate sizing to within 2 mm. In this series, we have used a relatively fixed length for the posterior ring, and revision or resizing has not been necessary. The posterior method may narrow the annulus to a greater degree than would occur with measured techniques; however, our echocardiographic data show no difference in transvalvular gradients between the annuloplasty groups. Unger-Graeber et al28 compared patients having a ring annuloplasty with others with no annuloplasty; although the mitral valve area was smaller in the annuloplasty group, there was no significant difference in transmitral gradient.28 These observations suggest that a better functional result without impaired hemodynamics can be obtained with a smaller annuloplasty.
Theoretically, the smaller annuloplasty may result in excess valvular tissue remaining within the left ventricle; if combined with a bulging septum, this may contribute to systolic anterior motion of the mitral valve and left ventricular outflow tract obstruction.30 31 We have not seen a high incidence of this complication.
Mitral incompetence invariably progresses.32 Rosen et al33 followed 31 patients with severe mitral regurgitation caused by prolapse who were relatively asymptomatic and found that 10.3% of patients per year deteriorate and develop symptoms that prompt valve repair or replacement. To continue observation in those with minimal symptoms and severe regurgitation exposes the patient to the risk of atrial fibrillation and worsening left ventricular function, the degree of which is often masked and potentially irreversible.
Degenerative mitral valvular disease is the most common pathological process seen, with prolapse and ruptured chordae present in nearly three fourths of patients. The patients are elderly (median age, 67 years), and close to half (48.8%) require additional surgical procedures. Despite this, operative mortality is extremely low (1.7%), and functional results are excellent in 95% of patients. These results and the simplicity of the posterior annuloplasty method support a more liberal attitude toward valve repair in minimally symptomatic or asymptomatic patients.6 34 35
| Acknowledgments |
|---|
| References |
|---|
|
|
|---|
This article has been cited by other articles:
![]() |
H. V. Schaff and R. M. Suri Multiple Valve Disease Card. Surg. Adult, January 1, 2008; 3(2008): 1129 - 1158. [Full Text] |
||||
![]() |
M. Arita, S. Tono, H. Kasegawa, and M. Umezu Multiple Purpose Simulator Using a Natural Porcine Mitral Valve Asian Cardiovasc Thorac Ann, December 1, 2004; 12(4): 350 - 356. [Abstract] [Full Text] [PDF] |
||||
![]() |
Y. K Mishra, H. Wasir, S. N Khanna, S. Shrivastava, Y. Mehta, and N. Trehan Multimodality Targeted Approach in Redo Off-Pump Coronary Artery Bypass Surgery Asian Cardiovasc Thorac Ann, March 1, 2003; 11(1): 7 - 10. [Abstract] [Full Text] [PDF] |
||||
![]() |
H. V. Schaff and D. H. Marsh Multiple Valve Disease Card. Surg. Adult, January 1, 2003; 2(2003): 1017 - 1045. [Full Text] |
||||
![]() |
M. Arita, H. Kasegawa, and M. Umezu Development of In-Vitro Evaluation System for Annuloplasty Rings Asian Cardiovasc Thorac Ann, March 1, 2001; 9(1): 14 - 18. [Abstract] [Full Text] [PDF] |
||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
Circulation Home | Subscriptions | Archives | Feedback | Authors | Help | AHA Journals Home | Search Copyright © 1995 American Heart Association, Inc. All rights reserved. Unauthorized use prohibited. |