Ann Thorac Surg 2003;76:487-492
© 2003 The Society of Thoracic Surgeons
Original article: cardiovascular
Bivalvular mechanical Mitral-Aortic valve replacement in 254 patients: Long-Term results—a 22-year follow-up
Jean-Paul Remadi, MDb*,
Olivier Baron, MDa,
C. Tribouilloy, PhDb,
Jean C. Roussel, MDa,b,
O. Al Habasch, MDa,
Philippe Despins, MDa,
Jean-Luc Michaud, MDa,
Daniel Duveau, MDa
a Cardiovascular Surgery Unit, G and R Laënnec University Hospital, Nantes, France
b Department of Cardiology, Amiens, France
Accepted for publication February 27, 2003.
* Address reprint requests to Dr Remadi, Cardiac Surgery Unit, Hôpital Sud, 80054 Amiens, France.
e-mail: remadi.jean-paul{at}chu-amiens.fr
 |
Abstract
|
|---|
BACKGROUND: We have retrospectively studied 254 patients who underwent a bivalvular mechanical mitral-aortic replacement in the cardiovascular and thoracic surgery unit of Nantes from 1979 to 1989. The follow-up was 22 years (1979 to 2001). The last patient was operated on 12 years before the end of the follow-up.
METHODS: All mitral prostheses were St. Jude Medical (SJM) bileaflet valves, and the aortic prostheses were 124 monodisc Björk-Shiley valves, 3 Sorin prostheses, and 127 St. Jude Medical bileaflet prostheses. The mean age was 56.8 ± 8.5 years with a sex ratio equal to 1. Rheumatism as the etiology predominated with 79.5%. Ninety-seven percent of the patients were followed for a total of 2,779 patient-years and a mean of 11.7 years.
RESULTS: Operative mortality was 7.08%. Freedom from overall mortality and valve-related mortality at 22 years were 45.7% ± 3.6% and 73.1% ± 3%, respectively. The linearized rates of thromboembolic and hemorrhagic events were 1.07% and 0.9% per patient-year, respectively. Multivariate analysis showed age (p < 0.002), sex (p < 0.01), and degenerative etiology (p = 0.04) as independent factors of late mortality, and age, sex, degenerative disease, and tricuspid pathology were related to valve-related mortality.
CONCLUSIONS: This study shows good results after mechanical mitral-aortic replacement in terms of survival rate and quality of life in surviving patients, and outlines the factors influencing long-term results as compared with isolated mitral valve replacement.
|
Introduction
|
|---|
Mitral-aortic valve replacement was first performed during the 1960s after the development of the first mechanical prostheses. These first generations of mechanical valves, combined with cardiopulmonary bypass techniques and as yet imperfect myocardial protection did not allow this combined surgery to be performed under optimal conditions for achieving satisfactory results. The advent of modern myocardial protection techniques and second-generation mechanical valves, led by the St. Jude Medical (SJM) prosthesis, transformed the short-term, medium-term, and long-term prognosis of these patients in the early 1980s.
We extracted a cohort of 254 patients undergoing double-valve replacement from a global series of 870 patients undergoing SJM mitral valve replacement (MVR) between 1979 and 1989 [1]. All of these 254 patients were therefore operated on by mechanical aortic valve replacement plus MVR, with an SJM valve in the mitral position. We conducted a retrospective study with very long-term follow-up (22 years), and a minimum follow-up of 12 years for the last patient operated on, to define the predictive factors of mortality for this double-valve replacement patient cohort.
|
Patients and Methods
|
|---|
We studied a cohort of 254 patients operated on for double mitral-aortic valve replacement in the cardiovascular and thoracic surgery unit of Nantes, between 1979 and 1989. These patients were extracted from a series of 870 patients undergoing MVR with an SJM valve during the same period. Therefore, these 254 patients represented 29.1% of the total patient cohort. The mean age of these patients was 56.8 ± 8.5 years with a sex ratio equal to 1 (Table 1). The predominant mitral valve disease was mitral insufficiency in 47% of cases (Table 1), whereas the predominant aortic disease was aortic insufficiency in 50% of cases. The largely predominant etiology was rheumatic heart disease (79.5%; Fig 1).
Patients were classified as New York Heart Association class III in 41% of cases (Fig 2).
Patients were only investigated by cardiac catheterization until 1984 and then mainly by echocardiography from 1985 onward. Left ventricular function was globally preserved, with a mean left ventricular ejection fraction of 0.58, and the pulmonary artery pressure was elevated with a mean value of 35 mm Hg. The cardiothoracic index calculated on the standard chest roentgenogram was increased, with a mean of 0.57 ± 0.2.
View larger version (58K):
[in this window]
[in a new window]
|
Fig 1. Causes of mitral and aortic valve disease. Numbers appearing next to each bar indicate number of patients. (CMO = obstructive cardiomyopathy.)
|
|
View larger version (62K):
[in this window]
[in a new window]
|
Fig 2. New York Heart Association class for patients undergoing double-valve replacement. Preoperative class (top) and class after 22 years of follow-up (bottom) is shown.
|
|
The standard procedure was performed by median sternotomy under cardiopulmonary bypass with moderate hypothermia at 28°C and intrapericardial cold liquid cooling (Shumway technique); a St. Thomas crystalloid cardioplegic technique was subsequently used from 1982 onward.
The mitral valve was accessed directly through the left atrium in Sondergaard’s groove in 82% of cases. Guiraudon’s right transatrial incision, through the roof of the left atrium, was never used in this series. The aortic valve was implanted after implantation of the mitral valve to facilitate exposure of the mitral ring. The most frequent diameter of the implanted valve was 21 mm in the aortic position and 29 mm in the mitral position (Fig 3).
View larger version (36K):
[in this window]
[in a new window]
|
Fig 3. Numbers of patients receiving various diameters of aortic (top) and mitral (bottom) prostheses. Numbers in labels for different prostheses indicate size in millimeters. (bs = Björk-Shiley; smj = St. Jude Medical.)
|
|
Mitral prostheses were SJM exclusively, whereas one half (n = 127) of the aortic valves were SJM, 124 consisted of a Björk-Shiley monostrut disk, and 3 were Sorin valves. No single factor was involved in the choice of prosthesis for the aortic position, and the implantations were consecutive. In the mitral position the SJM prosthesis had proved its hemodynamic superiority, thus all mitral implantations were performed with SJM valves. The associated operative procedures were as follows: 8 coronary artery bypass graft operations (3.1%), 25 tricuspid reconstructions (9.8%), and 2 tricuspid replacements (with bioprosthesis).
Follow-up
We updated an initial complete follow-up performed in 1994 by sending a letter to the patient, attending physician, and cardiologist. The nonresponse rate was 1.25% in 1994. The updated follow-up in September 2001 was performed by a patient telephone questionnaire, and the patient’s attending physician or cardiologist was also contacted. Eight patients were lost to follow-up in this updated survey, corresponding to a global follow-up rate of 96.85%. The cumulative follow-up was therefore 2,779 patient-years with a mean follow-up of 11.7 years.
The data were collected following the recommendations of the main international guidelines [2, 3].
A cohort of 440 patients operated on by isolated MVR with SJM prosthesis, extracted from the same cohort during the same period, whose follow-up was updated in 1998 [4], will be used for comparison in the comments.
Statistical methods
Continuous variables were expressed as mean ± standard deviation and compared using Student’s t test or Mann-Whitney U test, as appropriate. Qualitative variables were expressed as a percentage and compared by 
2 test or Fisher’s exact test, as appropriate. A p value less than 0.05 was considered to be significant. Multivariate analysis was performed to determine which factors were associated with deaths or events. Preoperative, intraoperative, and postoperative variables were studied by factorial analysis (multiple correspondence analysis). The variables were extracted from the model when their contribution to the formation of the factorial axis was greater than 80%. These extracted variables were submitted to a stepwise Cox multivariate analysis. A variable was identified as a significant independent factor when the p value was less than 0.05.
The Kaplan-Meier method was used for actuarial analysis. All late mortality and morbidity rates were expressed as linearized rates and by actuarial analysis. Linearized occurrence rates were calculated by dividing the observed number of occurrences of a particular event (death or complication) by the total number of years of patient follow-up. All operative deaths were included in actuarial survival analysis. The log-rank test was used for the comparison of the two event-free survival curves.
|
Results
|
|---|
Early results
The mean length of stay in intensive care was 28.74 ± 14.2 hours. The most frequent complication was low cardiac output syndrome (16.9%), responsible for a high operative mortality. Transient atrial fibrillation (AF) (18.5%) was the most frequent minor complication. On discharge from the hospital, after a mean hospital stay of 11.2 ± 5 days, the numbers of patients in sinus rhythm and AF were similar to those observed preoperatively. Figure 4
shows that the cardiac rhythm changed from sinus rhythm to AF or vice versa in an almost equivalent number of patients (almost equal to 7%), resulting in a stable number of patients in AF and in sinus rhythm. In fact, almost all patients with an altered rhythm during the postoperative phase regained their initial rhythm after the first month. One percent of patients developed complete atrioventricular block and required implantation of a permanent pacemaker.
View larger version (15K):
[in this window]
[in a new window]
|
Fig 4. Comparison of cardiac rhythm changes for all patients preoperatively and postoperatively, and as reported in initial follow-up (1994) and current follow-up (2001). (AF = atrial fibrillation; PM = pacemaker; SIN = sinus rhythm.)
|
|
The operative mortality (<30 days) was 7.08%. The leading cause of operative mortality was a low cardiac output syndrome (20%). Operative mortality was lower at the end of the series compared with the beginning of the series (7% versus 5%).
Univariate analysis demonstrated a significant correlation between early mortality and age more than 60 years (p < 0.03), left ventricular ejection fraction less than 0.40 (p < 0.01), degenerative disease (p < 0.01), and urgent procedure (p < 0.002). Multivariate analysis showed only age and urgent indication to be significantly correlated with operative mortality.
Late results
At the end of the follow-up, 124 patients were alive. The main causes of late overall mortality were extracardiac diseases (38.4%), predominantly neoplasms (Table 2).
Secondary heart failure was the main cause of cardiac deaths (75%), and the sudden ones were the main cause of valve-related deaths (Table 2; all the sudden ones were included in the valve-related deaths and no autopsies were performed). The freedom from overall mortality and valve-related deaths was 45.7% ± 3.6% and 73.1% ± 3%, respectively (Fig 5).
View larger version (17K):
[in this window]
[in a new window]
|
Fig 5. Actuarial survival curves for overall mortality (black symbols) and valve-related mortality (gray symbols). Patients at risk are shown below.
|
|
Univariate analysis demonstrated a significant correlation between overall mortality and age more than 60 years (p < 0.004), female sex (p < 0,01), mitral insufficiency (p < 0.001), and degenerative causes (p < 0.001). Multivariate analysis showed age (p < 0.0002) (Exp(coef)3.3), sex (p < 0.03) (Exp(coef)1.7), and degenerative causes (p < 0.04) (Exp(coef)1.25) to be significantly correlated with overall mortality, whereas age, sex, degenerative diseases, and tricuspid disease were correlated with valve-related mortality.
No cases of structural failure were reported. Linearized rates of thromboembolic accidents, valve thrombosis, and hemorrhagic accidents were 1.07%, 0.18%, and 0.9% per patient-year, respectively. The 22-year freedom from endocarditis and reoperation was 99.2% ± 1.6% and 93.4% ± 1.9%, respectively (Fig 6).
Paravalvular leak was the main indication for reoperation.
View larger version (18K):
[in this window]
[in a new window]
|
Fig 6. Valve-related morbidity as a function of reoperation (diamonds), thrombosis (gray squares), endocarditis (triangles), hemorrhage (black squares), paravalvular leak (small black rectangles), and thromboembolism (dashed line).
|
|
Clinical status of surviving patients
Among the 124 surviving patients, 49% were in sinus rhythm and 37% presented an AF (50% in sinus rhythm in 1994; Fig 4). Some patients required an implantation of a pacemaker some time after the operation: 14.3% of patients had a pacemaker at the last follow-up. The great majority (11.5% of the total of 14.3%) of patients subsequently requiring a pacemaker were previously in AF (Fig 4). The majority of patients were classified as New York Heart Association class I or II (96%; Fig 2). Therefore, 80% of patients considered that they had a normal quality of life, and 20% reported a less or very limited quality of life. The mean age of surviving patients at the last follow-up was 70 years. Twenty-five percent of patients were receiving too little or too much anticoagulant therapy (international normalized ratio < 2 or > 4, respectively) and often needed to adjust their treatment.
|
Comment
|
|---|
The 12-year follow-up for the last patient operated on in this series gives a 22-year survival of 45% with a reliable crude mortality rate of 55%.
The reliability of this actual survival is ensured by the mean follow-up of almost 15 years. The operative mortality in this series was 7.08%, which is included in the late mortality and therefore represents the main factor that needs to be improved to decrease this global mortality. This operative mortality rate is comparable to that reported in the literature [5–9]. The low cardiac output syndrome is the main cause of operative mortality, owing to the double-valve replacement procedure requiring a long operating time. The mean duration of cardiopulmonary bypass and aortic clamping times were more than 90 minutes and 120 minutes, respectively, comparable to the cardiopulmonary bypass and clamping times reported in the literature [5, 6]. The alteration of myocardial contractile function, even only transiently, can induce a postoperative low cardiac output syndrome that can be fatal. Myocardial protection is therefore essential during double-valve replacement to ensure satisfactory early postoperative results after one of the longest operative procedures in standard cardiac surgery.
In our series of patients operated on between 1979 and 1989, myocardial protection was improved as a result of the introduction of crystalloid cardioplegia in 1982. Introduction of a more sophisticated myocardial protection technique, together with an improvement of cardiopulmonary bypass techniques (oxygenator, membrane, biocompatible material, and so forth), has decreased the operative mortality of double-valve replacement. The current use of warm-blood or cold-blood cardioplegia further improves these immediate postoperative results with operative mortality rates often close to 5%. In our series, preoperative left ventricular function was significantly correlated with operative mortality by multivariate analysis and can be explained by the above analysis.
However, preoperative left ventricular function was not found to be a significant predictive factor for late mortality, as in the series of Turina and colleagues [9], but in their series all kinds of prostheses were used (bioprostheses and mechanical prostheses). Nevertheless, end-stage secondary heart failure is the main cardiac cause of late mortality. Therefore, left ventricular failure has to be managed correctly from the first symptoms to improve the short-term, medium-term, and long-term prognosis of these patients. The introduction of new treatments, such as angiotensin-converting enzyme inhibitors or angiotensin II inhibitors, has allowed long-term improvement of these forms of secondary left ventricular failure and, in particular, has improved the quality of life of some of these patients (15% in our series) with end-stage cardiac failure. No patient in our series underwent heart transplantation for end-stage cardiac failure, but this constitutes a useful treatment option for the management of the youngest patients with New York Heart Association class IV who are in cardiac failure. Severe tricuspid disease was found to be significantly related to valve-related deaths as in the series of Turina and associates [9]. Indeed these patients had "triple valvular" diseases with an important dilatation of the heart, and the operation was performed too late. If we compare the results of isolated MVR performed during the same period [4] with this series, we can outline two main factors that are predictors of late outcome: age more than 60 years and female sex, which are common for all cardiac surgery procedures. Degenerative causes are found to be related to late mortality only with bivalvular replacement. Because those patients often have a double regurgitation (aortic and mitral) with dilatation of the heart, dysfunction of the left ventricle occurs, and these patients therefore have a poor prognosis. The freedom from overall mortality at 10 and 19 years is comparable between isolated MVR and aortic valve replacement plus MVR (Fig 7).
If we compare the two actuarial curves, we can see that they are similar, other than at the beginning, owing to the operative mortality being higher for the bivalvular replacement, so the curve decreased quickly for the first year. The follow-up does not indicate any significant difference between the cohorts of patients who received a Björk-Shiley (124 patients) or SJM prosthesis (127 patients) implanted in the aortic position.
View larger version (22K):
[in this window]
[in a new window]
|
Fig 7. Actuarial survival curves comparing patients with bivalvular replacement (diamonds) and those with isolated mitral valve replacement (squares).
|
|
The leading causes of valve-related morbidity were sudden death and hemorrhagic accidents as already reported in our series of isolated MVR [4]. The great majority (90%) of hemorrhagic accidents (100% of fatal hemorrhagic accidents) were related to excessive anticoagulation, similar to patients undergoing MVR. The rate of patients with anticoagulation that was not adapted long term was relatively high at 25%. On the other hand, the anticoagulation-related events rate was relatively low, no doubt because of close control by the physician, after an immediate readjusting of the treatment in case of an important maladaptation of the international normalized ratio.
As the mean age of patients surviving at the end of follow-up was approximately 70 years, this lack of long-term adaptation of anticoagulation can be serious, as increasing age induces an increased risk of hemorrhage, but especially thromboembolism. Thus, Figure 6 shows that thromboembolism became predominant after the twelfth year: progressive dilatation of the heart and AF can explain this fact.
The proportion of surviving patients in sinus rhythm tended to increase, emphasizing the excess mortality in the population of patients with AF, although this factor was not found to be significantly correlated with late mortality. Therefore, even if there is an excessive mortality for those patients in AF, the difference in crude rates between surviving patients in sinus rhythm and patients with AF have to be interpreted in light of those patients fitted with a pacemaker. These patients had a satisfactory clinical status as well as a globally satisfactory quality of life for their age. The great majority of surviving patients considered that their quality of life was comparable to that of the age-matched population of individuals who were not operated on, similar to other series [6–8, 10].
The relatively high mitral paravalvular leak rate was without doubt explained by the suture procedure. Suture modification and the "expanding cuff" SJM prosthesis permitted, after 1984, such problems to be resolved.
The long-term results of double mechanical valve replacements are satisfactory in terms of both survival and quality of life, comparable to isolated MVR. The double-valve replacement provides a higher mortality rate but does not show any deterioration in long-term prognosis compared with isolated MVR.
Mitral-aortic valve diseases are now less frequent in industrialized countries as a result of the almost complete eradication of rheumatic fever. Degenerative causes have therefore become predominant. This surgical procedure now concerns a population of high-risk patients (degenerative causes, female sex, advanced age) that are all correlated significantly with higher late mortality. Therefore, when a procedure on the mitral valve is necessary in combination with an aortic valve replacement, mitral reconstruction remains preferable. When an MVR is necessary, preservation of the subvalvular apparatus is important, as the short-term, medium-term, and long-term prognoses are improved by the constant concern to preserve left ventricular function. Improved management of anticoagulant therapy also remains an important objective to further reduce morbidity and mortality of these patients. "Tri-valve" disease, with a tricuspid disease, has a poor long-term prognosis. Those patients must be operated on early, before left ventricular failure appears.
|
References
|
|---|
- Remadi J.P., Baron O., Bizouarn P., et al. Mitral valve replacement with the St. Jude Medical prosthesis: a 15-year follow-up. Ann Thorac Surg 1998;66:762-767.[Abstract/Free Full Text]
- Edmunds L.H., Jr, Clark R.E., Cohn L.H., Grunckemeir G.L., Miller D.C., Weisel R.D. Guidelines for reporting morbidity and mortality after cardiac valvular operations. Ann Thorac Surg 1996;62:932-935.[Abstract/Free Full Text]
- Gohlke-Barwolf C., Acar J., Burckhardt D., et al. Guidelines for prevention of thromboembolic events in valvular heart disease. Ad Hoc Committee of the Working Group on Valvular Heart Disease, European Society of Cardiology. J Heart Valve Dis 1993;2:398-410.[Medline]
- Remadi J.P., Baron O., Roussel J.C., et al. Isolated mitral valve replacement with St. Jude medical prosthesis: long term results: a follow-up of 19 years. Circulation 2001;103:1542-1545.[Abstract/Free Full Text]
- Arom K.V., Nicoloff D.M., Kersten T.E., Northrup W.F., III, Lindsay W.G., Emery R.W. Ten-year follow-up study of patients who had double valve replacement with the St. Jude Medical prosthesis. J Thorac Cardiovasc Surg 1989;98:1008-1016.[Abstract]
- Kratz J.M., Crawford F.A., Jr, Sade R.M., Crumbley A.J., Stroud M.R. St. Jude prosthesis for aortic and mitral valve replacement: a ten-year experience. Ann Thorac Surg 1993;56:462-468.[Abstract]
- John S., Ravikumar E., John C.N., Bashi V.V. 25-year experience with 456 combined mitral and aortic valve replacement for rheumatic heart disease. Ann Thorac Surg 2000;69:1167-1172.[Abstract/Free Full Text]
- Kinsley R., Antunes M., Colsen P. St. Jude valve replacement. J Thoracic Cardiovasc Surg 1986;92:349-360.[Abstract]
- Turina J., Stark T., Seifert B., Turina M. Predictors of the long term outcome after combined aortic and mitral valve surgery. Circulation 1999;100(Suppl 2):II-48-II-53.
- Mueller X.M., Tevaearai H.T., Strumpe F., et al. Long-term results of mitral-aortic valve operations. J Thorac Cardiovasc Surg 1998;115:1298-1309.[Abstract/Free Full Text]
This article has been cited by other articles:
|
|
|
|
 
M. Ricci, F. I. B. Macedo, M. R. Suarez, M. Brown, J. Alba, and T. A. Salerno
Multiple valve surgery with beating heart technique.
Ann. Thorac. Surg.,
February 1, 2009;
87(2):
527 - 531.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
S. Sawaki, A. Usui, T. Abe, M. Yoshikawa, T. Akita, and Y. Ueda
Late Mortality and Morbidity in Elderly Patients with Mechanical Heart Valves
Asian Cardiovasc Thorac Ann,
June 1, 2006;
14(3):
189 - 194.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
Determinants of operative mortality in valvular heart surgery.
J. Thorac. Cardiovasc. Surg.,
March 1, 2006;
131(3):
547 - 557.
|
|
|
Top
Abstract
Introduction
