| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Ann Thorac Surg 2004;78:1241-1247
© 2004 The Society of Thoracic Surgeons
a Department of Thoracic and Cardiovascular Surgery, University Hospital, Uppsala, Sweden
b Department of Cardiology, University Hospital, Uppsala, Sweden
c Department of Public Health Sciences, Karolinska Institute, Stockholm, Sweden
d IT Support Department, University of Uppsala, Uppsala, Sweden
Accepted for publication April 6, 2004.
* Address reprint requests to Dr Hellgren, Department of Thoracic and Cardiovascular Surgery, University Hospital, S-751 85 Uppsala, Sweden
laila.hellgren-johansson{at}akademiska.se
 |
Abstract |
|---|
 Top Abstract IntroductionMaterial and MethodsResultsCommentAcknowledgmentsReferences |
|---|
METHODS: Survival was analyzed in 784 patients undergoing mitral valve replacement from 1980 through 2000. Relative survival was estimated by relating the observed survival to the expected survival in a cohort, comparable regarding sex, age group, and calendar period, in the general Swedish population.
RESULTS: Early mortality was 9.7%. Relative survival rates after 5, 10, and 15 years were 83%, 70%, and 54%, respectively. The corresponding rates for observed survival were 75%, 56%, and 36%. Advanced New York Heart Association class, but not high age, was a risk factor for early mortality, whereas both factors decreased survival. An association between age and New York Heart Association class was found, the majority of old patients also being severely symptomatic. Patients in New York Heart Association classes I and II showed excellent relative survival.
CONCLUSIONS: Survival is reduced after mitral valve replacement in patients with severe preoperative symptoms, whereas patients with less-severe symptoms show excellent survival. Older patients were more often severely symptomatic at the time of surgery. Irrespective of age, surgery before the occurrence of advanced symptoms should improve the long-term outcome.
|
Introduction |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentAcknowledgmentsReferences |
|---|
The aim of this study was to investigate the impact of age and preoperative symptoms on survival after primary mitral valve replacement (MVR). In this study, relative survival was used as a measure of the excess mortality among patients undergoing mitral valve surgery compared with the mortality in the general population.
|
Material and Methods |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentAcknowledgmentsReferences |
|---|
There were 379 men (48%; mean age at operation, 63.5 years; range, 22 to 81 years) and 405 women (52%; mean age, 64.3 years; range, 18 to 83 years). One hundred twenty-eight patients (16%) had pure mitral stenosis, 510 patients (65%) had pure mitral insufficiency, and 146 patients (19%) had a combined lesion. These 784 patients constituted the study population. Coronary angiography was performed preoperatively in all patients 50 years or older and in all patients with angina or in whom coronary artery disease was suspected on a clinical basis. All operations were carried out with a standard technique for cardiopulmonary bypass and moderate hypothermia (25° to 32°C).
The policy was to recommend a bioprosthesis to patients 70 years of age or older. However, the type of prosthesis was left to the discretion of each surgeon. Of the 784 patients, 613 received a mechanical valve (St. Jude, 328 patients; Björk-Shiley, 191; Duromedics, 62; CarboMedics, 31; and other, 1 patient), and 171 patients received a bioprosthesis (Carpentier-Edwards, 55 patients; Sorin, 8; other tissue valve including St. Jude bovine valve, 108).
Patients with a mechanical valve were prescribed life-long anticoagulation treatment, whereas most patients with a bioprosthesis were recommended 3 months of this treatment, unless atrial fibrillation or other conditions indicated a high risk of thromboembolism.
Data Collection, Follow-Up, and Outcome Events
All clinical data were recorded prospectively and stored in a computer. New York Heart Association (NYHA) classification [18] of congestive heart failure was made on the basis of the preoperative assessment of the patient. A subdivision into NYHA class IIIA and IIIB was used, in which patients who suffered slight discomfort in their normal activities but were able to walk a mile on the flat at their own speed and could climb stairs slowly without undue discomfort were allocated to NYHA class IIIA.
A national registration number is allocated to every Swedish citizen. In January 2001, all patients were followed up with respect to survival by computerized linkage to two national registers based on this number, namely the Swedish Cause of Death Register [19] and a continuously updated population register. By use of these combined registers, all patients could be assigned a date of death or identified as being alive on December 31, 2000. The mean length of follow-up in the study group was 89.7 months. Death from any cause was the primary endpoint.
Statistical Methods
Logistic regression analysis [20] was used to identify factors related to early mortality (death from any cause within 30 days postoperatively). The odds ratio (OR) and corresponding 95% confidence intervals (CI) obtained from this analysis were used as a measure of the relative risk.
The Student's t test was used to compare differences in mean age among risk factors with dichotomous variables. One-way analysis of variance was used to compare differences in mean age among risk factors with categorical variables. Differences in mean age were considered significant if p was less than 0.05.
To test for differences in the distributions of risk factors in relation to NYHA classification, the Cochran-Armitage Trend Test was used [21]. Such differences were considered significant if p was less than 0.05.
The observed survival rate for long-term mortality (death from any cause after 30 days postoperatively) was calculated by the actuarial (life-table) method [22]. Risk factors associated with observed survival were identified in univariate and multivariate analyses based on the standard Cox proportional hazards model [23]. The relative hazard [exp(ßi)] and corresponding 95% CI were used as a measure of the risk of death in different groups. Separate models were applied to estimate relative hazards for the follow-up periods 0 to 5 years and more than 5 years. Stratified Cox models were used to explore any interaction between type of valve lesion and different risk factors. Finally, any interaction was tested by introducing an interaction variable.
The relative survival, as an estimate of the mortality related to MVR, was calculated as the ratio of observed to expected survival rate [24]. The expected survival rates were calculated from life tables compiled from the total population of Sweden stratified by sex, 5-year age group, and 5-year calendar period.
The observed/expected death ratio was calculated as the observed number of deaths in the study population related to the expected number of deaths in a cohort comparable with respect to sex, age, and intervention period in the general population.
The reference cohort in the general population was updated annually, with correction for withdrawals.
The following variables were entered into the logistic regression analyses and Cox analyses of early mortality and observed survival: demographic variables (age at operation, sex, year of surgery, previous myocardial infarction), symptoms and clinical status (dyspnea, CHF preoperatively, NYHA functional class, preoperative heart rhythm [sinus rhythm, atrial fibrillation, or other]), associated conditions (hypertension, diabetes, other serious diseases, eg, malignancies), preoperative catheterization data (presence of significant coronary artery disease [ie, with stenosis of 50% or more in at least one coronary artery]; type of lesion [pure mitral regurgitation, mitral stenosis, or combined lesion]), and characteristics of the surgical procedure (concomitant CABG, mechanical versus biologic prosthesis). The factors that gave significant information concerning observed survival in the univariate Cox analysis were used for analyses of relative survival.
The SAS 6.12 statistical program for PC (SAS Institute, Solna, Sweden) was used for the data processing and statistical analyses.
|
Results |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentAcknowledgmentsReferences |
|---|
70 years: OR, 2.1; 95% CI, 1.3 to 6.2), but did not influence early mortality when corrections were made for other risk factors.
|
|
|
In separate analyses, in which the effects of the identified risk factors were allowed to change between intervals, the effects of high age and advanced NYHA class were even stronger regarding deaths occurring more than 5 years after surgery (Table 2). In addition, there were no significant interaction effects between type of valve lesion and different risk factors, ie, the identified risk factors and effects of these factors were virtually the same in patients with mitral valve insufficiency, those with stenosis, and those with a combined lesion.
|
70 years; Fig 3) and 83% and 78%, respectively, in patients in the youngest group (<50 years). The patients in an advanced NYHA class, IIIB or IV, showed excess mortality, with a relative survival rate of 62% after 10 years (Fig 4). In contrast, patients with no or mild symptoms before surgery (NYHA classes I and II) had excellent relative survival of approximately 90% 5 to 15 years after surgery, ie, survival close to that in the corresponding group in the general population cohort.
|
|
|
|
Comment |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentAcknowledgmentsReferences |
|---|
Our study was based on consecutive patients within a defined geographic area. Selection was minimized, and the results should therefore be generally representative of patients undergoing this surgical procedure. The numbers of patients in the subgroups were sufficiently large to permit somewhat general conclusions to be drawn. With a mean follow-up of 7.5 years, corresponding to 5,296 patient-years, the study allows a comprehensive analysis of survival after mitral valve surgery.
In the total study population there was a distinct excess mortality as compared with Swedish population as a control cohort. During the 15-year follow-up of 784 patients, there were 384 total observed deaths, as compared to the expected number of 154.
Older age is often claimed to be associated with increased operative risks and to have a negative influence on the long-term outcome [1, 2, 15]. Because of these anticipated increased risks, nonsurgical treatment is more readily chosen in patients older than 75 years [1, 2]. In this study, however, older patients showed a moderately increased early risk, and after correction for severity of symptoms, high age was not an independent risk factor for early death. As expected, older patients had poorer long-term observed survival compared with younger patients. But after correction for the expected mortality within each age group, the relative survival in the oldest patients was worse compared with that in younger ones. Relative survival analyses provide more sophisticated information compared with a simple ratio of observed to expected deaths, especially when effects of age are in focus. As a result of the very small number of expected deaths in the youngest age group, a relatively small number of observed deaths will contradictorily result in a large observed/expected death ratio for these patients [25].
Notably, the majority of the patients older than 70 years were classified as NYHA IIIB or IV at the time of surgery. Irrespective of age, the preoperative symptomatic status was the most important risk factor for both early and late mortality. This study confirms previous findings that patients in less-advanced NYHA classes have low operative mortality and excellent long-term survival that is not different from that expected in the general population [12]. Consequently, not only in younger but also in older age groups, delayed surgery—resulting in progression of NYHA class symptoms—will expose the patients to increased early and late risks that might be avoidable with intervention before the occurrence of severe symptoms. However, the possibility of a lead-time bias must be considered. Patients operated on in an earlier and less symptomatic stage may have better survival as compared with more-symptomatic patients simply as a result of surgery at an earlier time point. The full answer to this question can only be obtained by a randomized trial in which the outcome with different strategies of surgical timing is compared in equally symptomatic patients.
It should be emphasized that in the present study only 6 patients were considered to be in NYHA class I, and thus no conclusions can be drawn regarding this group of patients. A surgical approach in patients with less pronounced symptoms is defensible only under certain conditions. First and foremost, a low operative risk is of importance. Also, the possibility of valve repair must be high. It must be kept in mind that valve replacement introduces the patient to a new disease process with the panorama of valve-related complications. In subgroups of patients with a high operative risk, the conservative approach appears to be reasonable until the symptoms become worse. The goal in most high-risk patients should be to consider surgery somewhere in the transition from NYHA class II to IIIA to ensure the most optimal short-term and long-term outcome.
Study Limitations
The present study was based on patients undergoing MVR, as too few patients had mitral valve repair during the study period to allow meaningful analyses. Today it is generally accepted that valve repair is associated with better survival as compared with replacement [1–2, 8, 13–15]. Even though valve repair is the preferred mode of correction, it is still not feasible in all patients. A substantial number of patients are still receiving a prosthesis. Information on the outcome and risk factors obtained from analyses of patients undergoing valve replacement is therefore important and may also to some extent be applicable to patients treated with mitral valve repair. There have been successive changes in surgical techniques for MVR, also, during the study period. Today chordal or valve-sparing techniques are more frequently used. Moreover, the management of patients undergoing mitral valve surgery has become more sophisticated, with improvements in preoperative medication, use of echocardiography, anesthesia, and postoperative care. In spite of these improvements, we did not find any significant increase in survival during the studied period.
In the present study, 20% of all deaths occurred within 30 days postoperatively. Exclusion of deaths within the first month biases the results in favor of surgery. There is therefore a risk that studies focusing on long-term survival rates, including the present one, will underestimate the excess mortality rates after cardiac operations. All early deaths must be regarded as excess mortality and should be taken into consideration when survival analyses are discussed. Most importantly, regardless of whether the rate of early mortality is low, the possibility of an early fatal outcome is still a major concern for each individual patient facing surgical intervention.
It is also a weakness of this study that NYHA classification had to be used as a surrogate measure for left ventricular function. Although it is subjective, it may at least theoretically reflect different aspects of left ventricular performance. The NYHA classification is widely used and forms the basis of current recommendations for surgery.
The patients included in this study differed in many respects, for example in their type of lesion—mitral valve insufficiency or stenosis. Each subgroup may have some specific risk factors or effect of such risk factors. However, all patients undergoing mitral valve surgery also have much in common, and many patients do not have clear-cut disease. One possible strategy for risk factor analyses is to perform separate models for each subgroup. In this study we chose an alternative approach, namely to identify interaction effects. No significant interaction effects were found; that is, in patients with different types of lesions or procedures the effects of risk factors, including NYHA class, were the same.
In summary, the present study clearly established that if MVR was performed before severe symptoms occurred, survival was excellent in all age groups. In contrast, all patients in NYHA class IIIB or IV showed distinct excess mortality. Elderly patients were more often severely symptomatic, but old age per se was not a risk factor for either early mortality or late excess mortality.
Conclusions
In conclusion, our findings indicate that all patients with a significant mitral valve lesion, irrespective of age, are likely to benefit from mitral valve surgery undertaken before advanced symptoms develop. All patients with a mitral valve lesion should undergo careful surveillance so that onset of symptoms or objective signs of left ventricular deterioration can be detected at an early stage.
|
Acknowledgments |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentAcknowledgmentsReferences |
|---|
|
References |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentAcknowledgmentsReferences |
|---|
hle E, Bergström R, Holmberg L, et al. Survival after coronary artery bypass grafting-experience from 4661 patients. Eur Heart J. 1994;15:1204–1211This article has been cited by other articles:
|
|
 |
|
  T. Gudbjartsson, T. Absi, and S. Aranki Mitral Valve Replacement Card. Surg. Adult, January 1, 2008; 3(2008): 1031 - 1068. [Full Text]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| ANN THORAC SURG | ASIAN CARDIOVASC THORAC ANN | EUR J CARDIOTHORAC SURG |
| J THORAC CARDIOVASC SURG | ICVTS | ALL CTSNet JOURNALS |
) and relative (•) survival after primary mitral valve replacement in patients who survived their first postoperative month (n = 708). The figure shows 95% confidence intervals at 5, 10, and 15 years and the numbers of patients at risk.
