HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Add to Personal Folders Download to citation manager Author home page(s): Eugene H. Blackstone Delos M. Cosgrove, III Paul Kerr Antonino Marullo Patrick M. McCarthy Bruce W. Lytle Permission Requests Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Gillinov, A. M. Articles by Lytle, B. W. Search for Related Content PubMed PubMed Citation Articles by Gillinov, A. M. Articles by Lytle, B. W. Related Collections Valve disease

J Thorac Cardiovasc Surg 2003;125:1372-1387
© 2003 The American Association for Thoracic Surgery

Surgery for Acquired Cardiovascular Disease

Mitral valve repair with aortic valve replacement is superior to double valve replacement

From the Department of Thoracic and Cardiovascular Surgery^a and the Department of Biostatistics and Epidemiology,^b The Cleveland Clinic Foundation, Cleveland, Ohio.

Received for publication Aug 20, 2002. Revisions requested Aug 28, 2002; revisions received Sept 9, 2002. Accepted for publication Sept 17, 2002. Address for reprints: A. Marc Gillinov, MD, The Cleveland Clinic Foundation, Department of Thoracic and Cardiovascular Surgery/F25, 9500 Euclid Ave, Cleveland, OH 44195. (E-mail: gillinom{at}ccf.org).

	Abstract

Top Abstract Introduction Patients and methods Results Discussion Appendix 1 Appendix: Discussion References

 
Objectives: Double valve replacement has been advocated for patients with combined aortic and mitral valve disease. This study investigated the alternative that, when feasible, mitral valve repair with aortic valve replacement is superior.
Patients and Methods: From 1975 to 1998, 813 patients underwent aortic valve replacement with either mitral valve replacement (n = 518) or mitral valve repair (n = 295). Mitral valve disease was rheumatic in 71% and degenerative in 20%. Mitral valve replacement was more common in patients with severe mitral stenosis (P = .0009), atrial fibrillation (P = .0006), and in patients receiving a mechanical aortic prosthesis (P = .0002). These differences were used for propensity-matched multivariable comparisons. Follow-up extended reliably to 16 years, mean 6.9 ± 5.9 years.
Results: Hospital mortality rate was 5.4% for mitral valve repair and 7.0% for replacement (P = .4). Survivals at 5, 10, and 15 years were 79%, 63%, and 46%, respectively, after mitral valve repair versus 72%, 52%, and 34%, respectively, after replacement (P = .01). Late survival was increased by mitral valve repair rather than replacement (P = .03) in all subsets of patients, including those with severe mitral valve stenosis. After repair of nonrheumatic mitral valves, 5-, 10-, and 15-year freedom from valve replacement was 91%, 88%, and 86%, respectively; in contrast, after repair of rheumatic valves, it was 97%, 89%, and 75% at these intervals.
Conclusions: In patients with double valve disease, aortic valve replacement and mitral valve repair (1) are feasible in many, (2) improve late survival rates, and (3) are the preferred strategy when mitral valve repair is possible.

	Introduction

Top Abstract Introduction Patients and methods Results Discussion Appendix 1 Appendix: Discussion References

 
Ten percent of patients with valvular heart disease have involvement of both aortic and mitral valves. ^1-4 Most groups advocate double valve replacement for this entity. ^5,6 However, some data suggest a survival advantage for the strategy of aortic valve replacement combined with mitral valve repair. ^7,8 The benefits of mitral valve repair in patients with isolated mitral valve disease are well documented. ^9-11 The purposes of this study of patients with double valve disease were to (1) identify features distinguishing patients having mitral valve repair versus replacement, (2) determine whether mitral valve repair confers a survival advantage over valve replacement, (3) identify which patients benefit most from mitral valve repair, and (4) assess the durability of the mitral valve repair.

	Patients and methods

Top Abstract Introduction Patients and methods Results Discussion Appendix 1 Appendix: Discussion References

 
Study group
From 1975 to 1998, 978 patients underwent simultaneous aortic and mitral valve surgery at The Cleveland Clinic Foundation. Of these, 813 underwent primary simultaneous aortic valve replacement and either mitral valve replacement (n = 518) or repair (n = 295) with or without concomitant tricuspid valve surgery or coronary artery bypass grafting. They were identified as follows. Initially, the Cardiovascular Information Registry (CVIR), whose data has been approved for use in research, and the Institutional Review Board were used to identify all patients having surgery for combined aortic and mitral valve disease. Medical records were then reviewed in detail to verify CVIR clinical data acquired concurrently with patient care. Patients with double valve repair or aortic valve repair and mitral valve replacement were excluded from the study. ¹² Patients undergoing double valve procedures for endocarditis were also excluded and are the subject of a separate report. ¹³

Definitions
Mitral valve disease was classified on the basis of analysis of clinical information, operative reports, catheterization reports, and echocardiograms. Degenerative mitral valve disease was considered to be present when the patient had mitral valve regurgitation resulting from leaflet prolapse and pathologic findings at operation consistent with degenerative disease. ¹⁴ Rheumatic mitral valve disease was considered to be present when the patient had mitral valve stenosis or pathologic and echocardiographic findings consistent with a rheumatic process. ¹⁵ Patients were deemed to have ischemic mitral valve regurgitation if they had papillary muscle infarction or mitral valve regurgitation caused by changes in left ventricular or anular geometry attributable to previous myocardial infarction. ¹⁶ Finally, patients with structurally normal mitral valves and no history of myocardial infarction were considered to have functional mitral valve regurgitation attributable to the effects of aortic valve disease.

Patient characteristics
Mean age of patients undergoing double valve replacement was 60 ± 13 years and among those receiving mitral valve repair was 61 ± 13 years (P = .16). Other patient characteristics, details of the cause and pathophysiology of the valve disease, cardiac and noncardiac comorbidity, and the operative procedure are given in Tables 1-3. Most patients had rheumatic valve disease and some degree of mitral valve stenosis. Forty-three percent of patients undergoing mitral valve repair had open mitral commissurotomy, and 59% had an annuloplasty. Aortic prostheses included 301 mechanical valves and 512 bioprostheses. Mitral valve prostheses included 202 mechanical valves and 317 bioprostheses. One patient had initial replacement with a mechanical prosthesis; a calcified anulus caused a perivalvar leak, and the valve was replaced with a bioprosthesis at the same operation. Operative reports did not permit the determination of the extent to which chordal-sparing techniques were used at mitral valve replacement. Concomitant procedures included coronary artery bypass grafting in 237 patients (29%), including 100 with mitral valve repair (34%) and 137 with mitral valve replacement (26%). Sixty-two patients (8%) had tricuspid valve repair.

Data analyses
Mitral valve replacement versus repair: Propensity analysis
Repair or replacement of the mitral valve was by surgeon choice. Therefore, attempts were made to adjust for selection factors in the analyses. The probability that the patient's mitral valve would be replaced rather than repaired was estimated by multivariable logistic regression by use of the demographic information, clinical status, cause of valve disease, pathophysiology of valve disease, and cardiac and noncardiac comorbidity variables that are listed in Appendix Table 1. In addition, because one of the possible selection criteria could have been whether the patient would receive long-term anticoagulation, the variable "mechanical aortic prosthesis" was included in the analysis. Variable selection for a parsimonious model was performed as described below under Multivariable Analysis.

Outcomes: Survival and durability of mitral valve repair
Time-related outcomes were all-cause death and reoperation. The analysis of reoperation focused primarily on mitral valve replacement after mitral valve repair. Nonparametric estimates used the Kaplan-Meier estimator. A parametric method was used to resolve the number of phases of instantaneous risk (hazard function) and to estimate their shaping parameters. ²⁰ (Available at http://clevelandclinic.org/heartcenter/hazard.)

Multivariable analyses
Potential risk factors were organized for entry into the various analyses as shown in Appendix Table 1. Exploratory analysis of these variables included correlation analysis, multiple stratified life-table analyses compared by use of the log-rank test, contingency table analyses, and simple t testing. Continuous and ordinal variables were assessed univariably by decile analysis to suggest transformations of scale to incorporate into the multivariable analyses to ensure that the relationship of these variables to outcome was well calibrated with respect to model assumptions.

For each of the hazard models, the multivariable analysis simultaneously incorporated variables into each hazard phase. For these, as well as the logistic regression analysis of propensity, a directed technique of entry of variables was used. ²¹ In all analyses of outcome, both the propensity score and the variable indicating mitral valve replacement rather than repair were always incorporated, regardless of their statistical significance. Additionally, interaction terms were formed between all variables and the indication of mitral valve replacement rather than repair to investigate the possibility that some variables had a differential influence in one or the other group.

The P value criterion for retention of variables in the final model was .1, except that both propensity score and the variable indicating mitral valve replacement rather than repair were retained in the analyses of outcome. This strategy does not, however, balance against type I and type II statistical error. Therefore, we supplemented the analyses with bootstrap random resampling, repeated 1000 times, to determine the likelihood of variables entering such an analysis at the 5% significance level. ^22,23

Analysis of benefit
To estimate the benefit of either mitral valve replacement or repair, the multivariable survival equation was solved for each patient twice, once as if the patient's mitral valve had been replaced and once again as if it had been repaired. The difference in predicted percent survival at 16 years for each of these two strategies was compared. The comparison included multiple linear regression analysis of the survival differences for their preoperative prediction as well as construction of stratified cumulative distribution curves.

Presentation
Mortality and survival estimates are accompanied by asymmetric 68% confidence limits (CL), comparable to ±1 standard error.

	Results

Top Abstract Introduction Patients and methods Results Discussion Appendix 1 Appendix: Discussion References

 
Mitral valve replacement versus repair: Propensity analysis
Patients were more likely to undergo mitral valve replacement if they had preoperative atrial fibrillation, severe mitral valve stenosis, calcified mitral leaflets, or if a mechanical aortic prosthesis was placed; mitral valve repair was more likely if there were morphologically normal valve leaflets with functional mitral valve regurgitation (Table 4). Earlier in the experience, valve replacement was more common (Appendix Figure 1, A).

View larger version (17K): [in this window] [in a new window]   Fig. 1. Trends across time in management of patients with double valve disease. In these graphs, each closed circle represents yearly proportion; a solid line is continuous probability by univariable logistic regression. A, Proportion of patients in whom mitral valve was replaced; B, proportion of patients coming to operation with rheumatic mitral valve disease versus degenerative mitral valve disease.

Survival advantage of mitral valve repair
Thirty-six (7.0%, CL 5.8% to 8.3%) patients died in the hospital after undergoing double valve replacement and 16 patients (5.4%, CL 4.1% to 7.1%) after aortic valve replacement and mitral valve repair (P = .4). The most common modes of death were heart failure (34 patients, 65%) and multiple organ dysfunction syndrome (10 patients, 19%).

There were 287 deaths after hospital discharge. After a double valve replacement, the survival rate was 93%, 86%, 72%, 52%, and 34% at 30 days, 1, 5, 10, and 15 years, respectively, after operation; after aortic valve replacement and mitral valve repair, the survival rate was 95%, 89%, 79%, 63%, and 46% at these same time intervals (P = .01, Figure 1, A). The instantaneous risk of death was highest immediately after operation and fell to its lowest level at 1 year, rising slowly thereafter. However, this late phase of hazard was consistently higher after mitral valve replacement than repair (Figure 1, B).

View larger version (24K): [in this window] [in a new window]   Fig. 1. Time-related death after double valve surgery. A, Survival. Each symbol represents a death, positioned according to the Kaplan-Meier estimator. Open squares represent patients undergoing aortic valve replacement and mitral valve repair, and open circles represent patients undergoing double valve replacement. Vertical bars are asymmetric confidence limits for these estimates. Superimposed are parametric survival estimates and their confidence limits (solid line and dashed line, respectively). Numbers in parentheses represent patients traced beyond the indicated interval. B, Instantaneous risk (hazard function) of death. Solid line (point estimates) and coarse dashed lines (confidence limits) represent patients undergoing double valve replacement (replace), and solid line enclosed within fine dashed lines represents patients undergoing aortic valve replacement and mitral valve repair (repair).

View larger version (29K): [in this window] [in a new window]   Fig. 2. Survival after double valve surgery stratified according to double valve replacement (replace) or aortic valve replacement and mitral valve repair (repair). Depiction is as in Figure 1. A, Rheumatic disease; B, nonrheumatic disease.

View larger version (15K): [in this window] [in a new window]   Fig. 3. Cumulative distribution of difference in percentage survival at 16 years between double valve replacement and aortic valve replacement and mitral valve repair, stratified according to rheumatic and nonrheumatic disease (see Analysis of Benefit under Patients and Methods). Only 1.7% of patients were predicted not to benefit from repair (negative portion of axis not shown).

View larger version (24K): [in this window] [in a new window]   Fig. 4. Survival after double valve surgery according to age at operation. Depiction is a nomogram from the multivariable analysis (Table 5) in which patient characteristics were entered as follows: NYHA Class II, left ventricular dysfunction less than grade 3, no important coronary disease, sinus rhythm, blood urea nitrogen 20 mg/dL, no kidney disease. A, Rheumatic disease; B, nonrheumatic disease.

View larger version (26K): [in this window] [in a new window]   Fig. 5. Freedom from mitral valve replacement after aortic valve replacement and mitral valve repair. Patients are stratified according to rheumatic or nonrheumatic disease. Depiction is as in Figure 1. A, Freedom; B, hazard function.

View larger version (19K): [in this window] [in a new window]   Fig. 6. Freedom from mitral valve reoperation after double valve surgery. Patients are stratified by both type of prosthesis used for mitral valve replacement and by cause if the mitral valve was repaired. Depiction is as in Figure 1, except that parametric confidence limits are suppressed for clarity.

	Discussion

Top Abstract Introduction Patients and methods Results Discussion Appendix 1 Appendix: Discussion References

During the last two decades, we have taken an aggressive approach to mitral valve repair. In spite of this, most patients with combined aortic and mitral valve disease have been managed with double valve replacement. This analysis identified certain characteristics associated with use of mitral valve repair rather than valve replacement. Patients requiring warfarin therapy for atrial fibrillation or a mechanical aortic prosthesis usually received a mechanical mitral prosthesis. Patients with severe mitral valve stenosis and mitral valve leaflet calcification were more likely to undergo mitral valve replacement rather than mitral commissurotomy. In contrast, patients with purely functional mitral valve regurgitation were most frequently treated by mitral valve repair, usually consisting of an annuloplasty alone.

This tendency to favor mitral valve replacement in certain groups of patients seemed logical. Patients requiring warfarin for other reasons might be expected to derive the greatest benefit from the excellent durability of a mechanical mitral valve prosthesis. Repair of calcified mitral valves presents a technical challenge with uncertain results, and patients with such valves can be treated expeditiously and reliably by mitral valve replacement. However, the long-term outcome data do not support these practices. None of the factors identified in the propensity analysis results in improved survival for patients having mitral valve replacement. Stated more explicitly, mitral valve replacement is associated with decreased long-term survival in all patients with double valve disease.

Survival advantage of mitral valve repair
The principal finding of this study is that mitral valve repair improves late survival in patients with double valve disease. This survival benefit extends to all patients, including the elderly, those with coronary artery disease, those with depressed left ventricular function, and those with rheumatic and nonrheumatic mitral valve disease.

Mitral valve repair in patients with single valve disease has been studied extensively. ^9-11,14,15 Since the introduction of standardized techniques for mitral valve reconstruction by Carpentier, ²⁴ Duran and Ubago, ²⁵ and others, mitral valve repair has become the surgical treatment of choice for mitral valve dysfunction. Numerous retrospective studies of patients with single valve disease have demonstrated important benefits of mitral valve repair over mitral valve replacement. ^9-11,14 Furthermore, it is likely that mitral valve repair confers a survival advantage in patients with single valve disease. ^9-11

Simultaneous operation on both valves is associated with a hospital mortality rate of 5% to 15%. ^26-28 Most contemporary series document a 10-year survival rate of 50% to 70% after double valve replacement; this is similar to late survival after isolated aortic or mitral valve replacement. ^1-4,7

There are little data examining the survival impact of mitral valve repair in patients with double valve disease. Kaul and coworkers ⁷ performed aortic valve replacement and mitral commissurotomy in 72 patients with rheumatic disease. They had no operative deaths and a 9-year survival rate in excess of 90%. Szentpetery ⁸ demonstrated the feasibility of aortic valve replacement and mitral valve repair in 38 patients, most of whom had degenerative mitral valve disease. With their small number of patients, they were unable to demonstrate a survival advantage to mitral valve repair. Mueller and colleagues ²⁹ found that patients having mitral valve repair and aortic valve replacement demonstrated a trend toward fewer valve-related complications than did patients having double valve replacement; however, they, too, reported similar survival rates after mitral valve repair and mitral valve replacement.

In the largest series of patients undergoing aortic valve replacement and mitral valve repair, Grossi and colleagues ⁵ analyzed 94 patients having this operative strategy. They were unable to demonstrate a survival advantage to mitral valve repair. Unfortunately, their study included a relatively small group of patients with limited follow-up, and they did not analyze all-cause death as an end point. In contrast, the current analysis of more than 5000 patient-years of follow-up demonstrates that mitral valve replacement is an independent predictor of late death in patients with double valve disease.

Survival advantage of mitral valve repair: Who benefits most?
Patients with double valve disease derive a survival benefit from mitral valve repair. Examination of different subgroups of patients revealed that patients with rheumatic mitral valve disease had a greater survival benefit with repair than did patients with nonrheumatic disease. This is accounted for in part by the finding that patients with rheumatic disease have better late survival than those with nonrheumatic disease, regardless of operative strategy. This may be attributable in part to preserved left ventricular function in patients with rheumatic mitral valve stenosis. Although the prevalence of rheumatic heart disease is declining in developed countries, most patients with double valve disease have rheumatic valves. From this study, we recommend that rheumatic mitral valves be repaired whenever possible in patients with double valve disease.

Durability of mitral valve repair
After repair of rheumatic mitral valves, 15-year freedom from valve replacement was 75%; in contrast, 15-year freedom from valve replacement was 86% after repair of nonrheumatic valves. Rheumatic cause decreases the durability of mitral valve repair. However, the durability of repaired rheumatic mitral valves exceeds that of bioprostheses in the mitral position. Most importantly, rheumatic valve repair confers a survival advantage in spite of somewhat limited 15-year durability.

In most series of patients with single valve disease, 40% to 75% of rheumatic mitral valves can be repaired. ²⁴ Durability of mitral valve repair in the rheumatic population is lower than that in patients with degenerative disease. ^14,24 After repair for rheumatic mitral valve regurgitation, 15-year freedom from reoperation is 76%. ²⁴ After open mitral valve commissurotomy for mitral valve stenosis, 10-year freedom from reoperation is 80% to 90%. ^15,30-32 Factors reducing the durability of mitral valve repair in rheumatic disease include younger patient age, pure mitral valve regurgitation, mixed mitral valve regurgitation and mitral valve stenosis, and leaflet calcification. ^15,30-32 However, late survival is excellent after repair of rheumatic mitral valves, and most authorities favor repair in such patients when feasible. ^30-32 Although late mitral valve reoperation for failed repair may be challenging in patients with an aortic prosthesis, current data support an initial strategy of mitral valve repair in patients with rheumatic double valve disease when repair is feasible.

Limitations
This is a nonrandomized clinical study. Using the propensity score, we have attempted to adjust the multivariable analyses of outcomes for nonrandom selection factors related to choice of valvar procedure. However, it can not adjust for unmeasured variables. The decision to repair or replace the mitral valve was made by the surgeon. Not all mitral valves are amenable to repair, particularly those with extensive leaflet calcification. However, none of the surgeons in retrospect were able to identify with certainty which of these patients they would not consider eligible for at least attempted repair. Changing at times across the span of this study, opinions differed and evolved about eligibility for repair. Thus, we have included in the study all patients who underwent operation during the entire time frame.

Serial echocardiographic follow-up assessment of mitral valve function was unavailable in most patients undergoing mitral valve repair. Therefore, the data did not allow identification of patients who had recurrent mitral valve dysfunction but did not undergo reoperation, precluding an assessment of durability on the basis of both reoperation and recurrent valve dysfunction.

The outcomes analyzed were death and mitral valve reoperation. Occurrence of valve-related complications (thromboembolism, endocarditis, anticoagulant-related hemorrhage) have been well studied and reported but were not central to the purposes of this study. We were unable to analyze the impact of chordal preservation on results after mitral valve replacement.

Clinical inferences and decision making
Mitral valve repair is possible in many patients with double valve disease. It improves late survival rates and is more durable than a bioprosthesis. Consequently, when a mitral valve that is amenable to repair is encountered in a patient with double valve disease, it should be repaired rather than replaced to ensure the best long-term outcome.

	Appendix 1

Top Abstract Introduction Patients and methods Results Discussion Appendix 1 Appendix: Discussion References

 

	Appendix: Discussion

Top Abstract Introduction Patients and methods Results Discussion Appendix 1 Appendix: Discussion References

 
Dr Cary W. Akins (Boston, Mass). I congratulate Dr Gillinov for his excellent presentation of this complex analysis of mitral valve repair versus replacement when performed in combination with aortic valve replacement. Unfortunately, I did not have the manuscript long enough to fully evaluate the complex statistical methods, even if I could have understood them, so I will try to look at some of the basic issues concerning the study.

All surgical groups, including our own, who have reported the advantages of mitral valve reconstruction versus replacement in single valve disease have the a priori notion that the conclusion of this manuscript must be correct. My innate prejudice favoring mitral valve repair makes me believe the conclusion, but I am not sure that the study actually proves the fact.

The study is retrospective and not randomized. The two patient groups are different in significant ways. Although many factors were assessed to try to statistically determine why surgeons chose one operation over another, there are important areas of missing information about factors that have been used to calculate propensities and results.

For example, in the manuscript, for the 813 patients studied, information on mitral valve pathophysiology, that is, stenosis or regurgitation, was available for less than two thirds of the patients, yet mitral valve stenosis is listed as a significant predictor of the operation performed. Also, there are no data in the study on two factors that are known to impact long-term survival in mitral valve disease, namely the severity of the mitral valve disease as represented by levels of pulmonary hypertension or a history of stroke, nor is the history of malignancy evaluated.

Many factors that lead surgeons who know their patients well to choose one operation over another are often not well documented in a hospital chart, let alone quantifiable. It is my contention that surgeons can frequently judge which patients have limited potential for long-term survival exclusive of their valvular disease and will perform the operation that is most expedient. To then attribute late all-cause death to the choice of operation seems inappropriate. Even Dr. Blackstone's extraordinary statistical powers can not accommodate for missing data or truly account for surgeon bias.

There is another equally important feature of this nonprospective, nonrandomized trial that merits attention. The implication in this comparative study is that mitral valve repair was an option for all patients. I doubt that is true. Even in the hands of most experienced surgeons, end-stage calcific mitral valve disease is not amenable to reconstruction. The hypothesis of the study would be easier to validate if the two patient groups studied both contained only patients for whom repair was a viable option.

In terms of the operation performed, the authors do not discuss the issue of chordal-sparing mitral valve replacement. Had this technique been used with mitral valve replacement, do you believe that there would have been better long-term results, as some authors contend?

The authors also state that patients with rheumatic disease have better survival rates than those with nonrheumatic disease and suggest that it is due to better preserved ventricular function, but they did not relate this to the presence of concomitant coronary artery disease. Could concomitant coronary artery disease be the reason for your findings?

I again congratulate Dr. Gillinov and his colleagues at the Cleveland Clinic on this provocative study, and, although I want to believe the inference, I am not sure that it has been proven.

Dr Gillinov. I will try to address all those points one by one, although this may take a while. You are absolutely correct, this is a retrospective clinical study. It is not a randomized prospective study. Such a study would be virtually impossible to perform, and, therefore, we have to analyze the data that are available, and we have done our best to do that. You do note that we have not reported the pathophysiology of mitral valve disease in about one third of the patients. The patients for whom we have specified the pathophysiology are those who had either preoperative or intraoperative echocardiography. We are now going back to gather the data that we can from catheterization reports, but we believe that the data from echocardiography is more reliable because we can more accurately judge the degree of mitral stenosis or regurgitation in these patients. We will be supplementing the data with catheterization reports.

We have not included the variable pulmonary hypertension. I believe that is a good suggestion, and it would be valuable to go back and gather that data and to determine the impact of preoperative pulmonary hypertension on the outcome of these patients.

As far as comorbid conditions such as stroke and malignancy, we do not make our decisions for the valvular procedure in a vacuum. We try to choose the best procedure that applies to the patient, and certainly a patient who has an extremely limited life expectancy on the basis of a progressive cancer might get a different operation from a 40-year-old individual. Again, though, those sorts of variables, as you correctly state, are difficult to capture in this sort of analysis.

We do not attribute all of the deaths in this study to the choice of valve procedure. The people who died are dead. It really doesn't matter of what they died. However, we did find an important relationship between the choice of valvular procedure and the mortality rate, and we believe that this is, in fact, a true relationship. We also agree, as everyone knows, mitral valve repair is not an option for all patients. There are valves that can not be repaired. We would temper our conclusions by saying mitral valve repair should be considered in all patients. If it is a heavily calcified rheumatic mitral valve with a completely fused subvalvular apparatus, this valve probably can not be repaired and should be replaced; and, when doing the replacement, on the basis of data accumulated by Dr David and others, we currently do spare as much of the subvalvular apparatus as possible. When looking for this data in old operative reports, we were unable to find it in most patients, and, therefore, we believed that it would be incorrect to try to analyze an incomplete data set.

Finally, with regard to your question of could the better survival rate in patients with rheumatic disease be related to the prevalence of coronary artery disease in these patients, the answer is no. We did correct for the presence of coronary artery disease. The patients with rheumatic disease simply lived longer.

Dr Tirone E. David (Toronto, Ontario, Canada). If you take a look at the clinical outcomes of mitral valve surgery, aside from things we know like age, New York functional class, and ventricular function before disease, the pathologic process, unlike in the aortic valve, plays a major role in long-term survival. Ischemic mitral valve disease is notoriously a bad disease. If you have severe mitral valve regurgitation caused by ischemia, your patient won't be alive for very many years regardless of whether you do repair or replacement. Conversely, if you have a myxomatous mitral valve, they do very well; they do very well with mitral valve replacement or repair. Did you look at the pathologic process and the effect it had on survival? Was it different in repair or replacement?

Dr Gillinov. Yes, all patients benefited from repair versus replacement. The incremental benefit of repair was greatest in those with rheumatic disease, but it was still present in those with degenerative disease. Only 1% of the patients in this study had ischemic disease, so we did not have an adequate number of patients to comment on patients who have ischemic mitral valve disease and concomitant aortic valve disease.

Dr David. Yet one of your graphs showed that survival with mitral valve replacement with a mechanical valve is almost identical to repair of all causes, telling me perhaps isn't the mortality rate largely related to a failed bioprosthesis or the reoperation it required?

Dr Gillinov. The slide that had the mechanical valves on it was freedom from reoperation. The choice of prosthesis when the mitral valve was replaced did not influence survival rates, but it did influence durability.

Dr Eugene A. Grossi (New York, NY). I have a question about freedom from reoperation in the rheumatic group. Was there a difference in freedom from reoperation between those patients who had mitral stenosis and primarily had a commissurotomy versus mitral valve reconstruction for those who had rheumatic insufficiency?

Dr Gillinov. We did not find any risk factors for reoperation in the patients with rheumatic mitral valve disease.

	Acknowledgments

 
We thank Drs Nicholas G. Smedira and Joseph F. Sabik for inclusion of their patients in this study; Colleen Vahcic, Deborah Gladish, and the other members of the thoracic and cardiovascular research team for their efforts in assembling and verifying the clinical data and performing the follow-up; John Hendricks and Linda DiPaola for constructing the data set; and, Tess Knerik for editorial assistance.

	Footnotes

 
Read at the Eightieth Annual Meeting of The American Association for Thoracic Surgery, Toronto, Ontario, Canada, April 30-May 3, 2000.

	References

Top Abstract Introduction Patients and methods Results Discussion Appendix 1 Appendix: Discussion References

 

1. Nakano K, Koyanagi H, Hashimoto A, Kitamura M, Endo M, Nagashima M, et al. Twelve years' experience with the St Jude Medical valve prosthesis. Ann Thorac Surg. 1994;57:697-703.[Abstract]
   
2. Baudet EM, Puel V, McBride JT, Grimaud JP, Roques F, Clerc F, et al. Long-term results of valve replacement with the St Jude Medical prosthesis. J Thorac Cardiovasc Surg. 1995;109:858-70.[Abstract]
   
3. Kinsley RH, Antunes MJ, Colsen PR. St Jude Medical valve replacement: an evaluation of valve performance. J Thorac Cardiovasc Surg. 1986;92:349-60.[Abstract]
   
4. Hartz RS, Fisher EB, Finkelmeier B, DeBoer A, Sanders JH Jr, Moran JM, et al. An eight-year experience with porcine bioprosthetic cardiac valves. J Thorac Cardiovasc Surg. 1986;91:910-7.[Abstract]
   
5. Grossi EA, Galloway AC, Miller JS, Ribakove GH, Culliford AT, Esposito R, et al. Valve repair versus replacement for mitral insufficiency: when is a mechanical valve still indicated? J Thorac Cardiovasc Surg. 1998;115:389-96.[Abstract/Free Full Text]
   
6. Brown PS Jr, Roberts CS, McIntosh CL, Swain JA, Clark RE. Relation between choice of prostheses and late outcome in double-valve replacement. Ann Thorac Surg. 1993;55:631-40.[Abstract]
   
7. Kaul P, John S, John CN, Bashi VV, Ravikumar E, Choudhry U, et al. Aortic valve replacement with concomitant open mitral valvotomy: early results and long term follow-up in 72 consecutive patients. Indian Heart J. 1993;45:113-5.[Medline]
   
8. Szentpetery S, Rich JB, Azar H, Newton JR, Tenzer MM. Mitral valve repair combined with aortic valve replacement. J Heart Valve Dis. 1997;6:32-6.[Medline]
   
9. Perier P, Deloche A, Chauvaud S, Fabiani JN, Rossant P, Bessou JP, et al. Comparative evaluation of mitral valve repair and replacement with Starr, Björk, and porcine valve prostheses. Circulation. 1984;70(Suppl):I187-92.
   
10. Enriquez-Sarano M, Schaff HV, Orszulak TA, Tajik AJ, Bailey KR, Frye RL. Valve repair improves the outcome of surgery for mitral regurgitation: a multivariate analysis. Circulation. 1995;91:1022-8.[Abstract/Free Full Text]
    
11. Sand ME, Naftel DC, Blackstone EH, Kirklin JW, Karp RB. A comparison of repair and replacement for mitral valve incompetence. J Thorac Cardiovasc Surg. 1987;94:208-19.[Abstract]
    
12. Gillinov AM, Blackstone EH, White J, Howard M, Ahkrass R, Marullo A, et al. Durability of combined aortic and mitral valve repair. Ann Thorac Surg. 2001;72:20-7.[Abstract/Free Full Text]
    
13. Gillinov AM, Diaz R, Blackstone EH, Pettersson GB, Sabik JF, Lytle BW. Double valve endocarditis. Ann Thorac Surg. 2001;71:1874-9.[Abstract/Free Full Text]
    
14. Gillinov AM, Cosgrove DM, Blackstone EH, Diaz R, Arnold JH, Lytle BW, et al. Durability of mitral valve repair for degenerative disease. J Thorac Cardiovasc Surg. 1998;116:734-43.[Abstract/Free Full Text]
    
15. Nakano S, Kawashima Y, Hirose H, Matsuda H, Shirakura R, Sato S, et al. Reconsiderations of indications for open mitral commissurotomy based on pathologic features of the stenosed mitral valve. J Thorac Cardiovasc Surg. 1987;94:336-42.[Abstract]
    
16. Gillinov AM, Wierup PN, Blackstone EH, Bishay ES, Cosgrove DM, White J, et al. Is repair preferable to replacement for ischemic mitral regurgitation? J Thorac Cardiovasc Surg. 2001;122:1125-41.[Abstract/Free Full Text]
    
17. Rosenbaum P, Rubin DB. The central role of the propensity score in observational studies for causal effects. Biometrika. 1983;70:41-55.[Abstract/Free Full Text]
    
18. Cook EF, Goldman L. Performance of tests of significance based on stratification by a multivariate confounding score or by a propensity score. J Clin Epidemiol. 1989;42:312-24.
    
19. Blackstone EH. Comparing apples and oranges. J Thorac Cardiovasc Surg. 2002;123:8-15.[Free Full Text]
    
20. Blackstone EH, Naftel DC, Turner ME Jr. The decomposition of time-varying hazard into phases, each incorporating a separate stream of concomitant information. J Am Stat Assoc. 1986;81:615-24.
    
21. Baskerville JC, Toogood JH. Guided regression modeling for prediction and exploration of structure with many explanatory variables. Technometrics. 1982;24:9-17.
    
22. Breiman L. Bagging predictors. Machine Learning. 1996;26:123-40.
    
23. Blackstone EH. Breaking down barriers: helpful breakthrough statistical methods you need to understand better. J Thorac Cardiovasc Surg. 2001;122:430-9.[Free Full Text]
    
24. Deloche A, Jebara VA, Relland JYM, Chauvaud S, Fabiani JN, Perier P, et al. Valve repair with Carpentier techniques: the second decade. J Thorac Cardiovasc Surg. 1990;99:990-1002.[Abstract]
    
25. Duran CG, Ubago JLM. Clinical and hemodynamic performance of a totally flexible prosthetic ring for atrioventricular valve reconstruction. Ann Thorac Surg. 1976;22:458-63.[Abstract]
    
26. Arom KV, Nicoloff DM, Kersten TE, Northrup WF III, Lindsay WG, Emery RW. Ten-year follow-up study of patients who had double valve replacement with the St Jude Medical prosthesis. J Thorac Cardiovasc Surg. 1989;98:1008-16.[Abstract]
    
27. Bortolotti U, Milano A, Thiene G, Guerra F, Mazzucco A, Talenti E, et al. Long-term durability of the Hancock porcine bioprosthesis following combined mitral and aortic valve replacement: an 11-year experience. Ann Thorac Surg. 1987;44:139-44.[Abstract]
    
28. Bortolotti U, Milano A, Testolin L, Tursi V, Mazzucco A, Gallucci V. Influence of type of prosthesis on late results after combined mitral-aortic valve replacement. Ann Thorac Surg. 1991;52:84-91.[Abstract]
    
29. Mueller XM, Tevaearai HT, Stumpe F, Fischer AP, Hurni M, Ruchat P, et al. Long-term results of mitral-aortic valve operations. J Thorac Cardiovasc Surg. 1998;115:1298-309.[Abstract/Free Full Text]
    
30. Eguaras MG, Luque I, Montero A, Garcia MA, Calleja F, Roman M, et al. A comparison of repair and replacement for mitral stenosis with partially calcified valve. J Thorac Cardiovasc Surg. 1990;100:161-6.[Abstract]
    
31. Gometza B, Al-Halees Z, Shahid M, Hatle LK, Duran CMG. Surgery for rheumatic mitral regurgitation in patients below twenty years of age: an analysis of failures. J Heart Valve Dis. 1996;5:294-301.[Medline]
    
32. David TE. Editorial: the appropriateness of mitral valve repair for rheumatic mitral valve disease. J Heart Valve Dis. 1997;6:373-4.[Medline]
    

This article has been cited by other articles:

				2006 WRITING COMMITTEE MEMBERS, R. O. Bonow, B. A. Carabello, K. Chatterjee, A. C. de Leon Jr, D. P. Faxon, M. D. Freed, W. H. Gaasch, B. W. Lytle, R. A. Nishimura, et al. 2008 Focused Update Incorporated Into the ACC/AHA 2006 Guidelines for the Management of Patients With Valvular Heart Disease: A Report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines (Writing Committee to Revise the 1998 Guidelines for the Management of Patients With Valvular Heart Disease): Endorsed by the Society of Cardiovascular Anesthesiologists, Society for Cardiovascular Angiography and Interventions, and Society of Thoracic Surgeons Circulation, October 7, 2008; 118(15): e523 - e661. [Full Text] [PDF]

				R. O. Bonow, B. A. Carabello, K. Chatterjee, A. C. de Leon Jr, D. P. Faxon, M. D. Freed, W. H. Gaasch, B. W. Lytle, R. A. Nishimura, P. T. O'Gara, et al. 2008 Focused Update Incorporated Into the ACC/AHA 2006 Guidelines for the Management of Patients With Valvular Heart Disease: A Report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines (Writing Committee to Revise the 1998 Guidelines for the Management of Patients With Valvular Heart Disease) Endorsed by the Society of Cardiovascular Anesthesiologists, Society for Cardiovascular Angiography and Interventions, and Society of Thoracic Surgeons J. Am. Coll. Cardiol., September 23, 2008; 52(13): e1 - e142. [Full Text] [PDF]

				P. W.M. Fedak, P. M. McCarthy, and R. O. Bonow Evolving Concepts and Technologies in Mitral Valve Repair Circulation, February 19, 2008; 117(7): 963 - 974. [Full Text] [PDF]

				S. Talwar, A. Mathur, S. K. Choudhary, R. Singh, and A. S. Kumar Aortic Valve Replacement With Mitral Valve Repair Compared With Combined Aortic and Mitral Valve Replacement Ann. Thorac. Surg., October 1, 2007; 84(4): 1219 - 1225. [Abstract] [Full Text] [PDF]

				M. Nakajima, K. Tsuchiya, Y. Okamoto, and F. Suetsugu Intraoperative saline injection leak test a simple method to assess mitral valve repair when a simultaneous aortotomy does not allow pressurization of the left ventricle Interactive CardioVascular and Thoracic Surgery, October 1, 2007; 6(5): 580 - 581. [Abstract] [Full Text] [PDF]

				K. Kuwaki, N. Kawaharada, K. Morishita, T. Koyanagi, H. Osawa, T. Maeda, and T. Higami Mitral Valve Repair Versus Replacement in Simultaneous Mitral and Aortic Valve Surgery for Rheumatic Disease Ann. Thorac. Surg., February 1, 2007; 83(2): 558 - 563. [Abstract] [Full Text] [PDF]

				S M Tuladhar and P P Punjabi Surgical reconstruction of the mitral valve Heart, October 1, 2006; 92(10): 1373 - 1377. [Abstract] [Full Text] [PDF]

R. O. Bonow, B. A. Carabello, K. Chatterjee, A. C. de Leon Jr, D. P. Faxon, M. D. Freed, W. H. Gaasch, B. W. Lytle, R. A. Nishimura, P. T. O'Gara, et al. ACC/AHA 2006 Guidelines for the Management of Patients With Valvular Heart Disease: A Report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines (Writing Committee to Revise the 1998 Guidelines for the Management of Patients With Valvular Heart Disease) Developed in Collaboration With the Society of Cardiovascular Anesthesiologists Endorsed by the Society for Cardiovascular Angiography and Interventions and the Society of Thoracic Surgeons J. Am. Coll. Cardiol., August 1, 2006; 48(3): e1 - e148. [Full Text] [PDF]