- Email: [email protected]
Selecting patients with mitral regurgitation and left ventricular dysfunction for isolated mitral valve surgery
CARDIOVASCULAR
Selecting Patients With Mitral Regurgitation and Left Ventricular Dysfunction for Isolated Mitral Valve Surgery Constance K. Haan, MD, Cristina I. Cabral, MD, Donald A. Conetta, MD, Laura P. Coombs, PhD, and Fred H. Edwards, MD Divisions of Cardiothoracic Surgery and Cardiology, University of Florida, Jacksonville, Florida, and The Outcomes Research and Assessment Group, Duke Clinical Research Institute, Durham, North Carolina
Background. American College of Cardiology/ American Heart Association (ACC/AHA) Guidelines state that patients with an ejection fraction (EF) of 30% or less should not undergo mitral valve replacement for mitral regurgitation (MR). We sought to establish, using a national cardiac surgery database, whether patients with left ventricular dysfunction may safely undergo mitral valve surgery for MR, and if so, which ones. Methods. We queried the Society of Thoracic Surgeons (STS) National Database to identify patients who had isolated mitral valve replacement or repair for MR between 1998 and 2001. Mortality and morbidity outcomes were compared by EF category (< 30% vs > 30%), and observed mortality compared by EF group, stratified by predicted risk for mortality. A classification and regression tree (CART) model was then used to determine which patient characteristics contributed most to designate the high-risk patient. Results. Of the 14,582 patients who had mitral valve
surgery, 727 had an EF of 30% or less and 13,855 had an EF of more than 30%. Observed mortality rates were higher for patients with an EF of 30% or less (5.4% vs 3.1%). However, for low-risk to medium-risk patients, mortality rates remained fairly constant across levels of EF. Mortality is notably increased in the high-risk patients (predicted risk > 10%). A classification tree identifies three key characteristics for high risk: age more than 75 years, renal failure, and emergent or salvage procedure. Conclusions. When the predicted mortality risk is less than 10%, EF has minimal impact on operative mortality for mitral regurgitation. In contrast to the ACC/AHA Guidelines, our data show that operative risk for mitral valve surgery is not prohibitive for most patients with ventricular dysfunction.
C
Accepted for publication April 1, 2004.
recommend, “if mitral valve repair appears likely, surgery should still be contemplated, provided ejection fraction is more than 30%.” [12] By contrast, some recent studies show that mitral valve surgery offers symptomatic improvement and survival benefit to MR patients with poor LV function [13–19]. Further, evidence indicates that mitral procedures in patients with MR and low EFs may be done with acceptable outcomes [13, 15, 20 –23]. It is known that operative outcomes associated with mitral valve repair are different from those associated with mitral valve replacement surgery [24 –26]. It is unclear how these outcome differences are influenced by patient variability and selection, or patient or surgeon bias for choice of operation, or both. Furthermore, most reports deal with small, and relatively select patient populations—with study patient populations ranging in size from 9 to 478 [2–10, 13–26]. We used a national database to study the inherently larger available number of patients to address the following questions:
Address reprint requests to Dr Haan, Division of Cardiothoracic Surgery, The Cardiovascular Center, UFHSC/J, 655 W Eighth St, Jacksonville, FL 32209; e-mail: [email protected].
1. Should mitral valve surgery be offered to patients with significant LV dysfunction?
ardiac surgery for mitral regurgitation (MR) has long been aimed toward intervening before significant deterioration of left ventricular (LV) function occurs [1]. This is intended to minimize perioperative morbidity and mortality and to optimize recovery and postoperative function [2– 4]. Mitral valve surgery in patients with MR and severe LV dysfunction has been associated with high operative mortality [5– 8]. Postoperative LV function and survival have been shown to be associated with preoperative LV function [6, 9, 10]. The firm recommendation has been made that MR patients with an ejection fraction (EF) of less than 30% should not be offered mitral valve surgery [11]. In fact, the ACC/AHA Guidelines for the Management of Patients with Valvular Heart Disease state that patients with an EF of less than 30% should not undergo mitral valve replacement for chronic MR, especially if the mitral valve apparatus [12] can not be preserved. Furthermore, the ACC/AHA Guidelines for Valve Surgery
© 2004 by The Society of Thoracic Surgeons Published by Elsevier Inc
(Ann Thorac Surg 2004;78:820 –5) © 2004 by The Society of Thoracic Surgeons
0003-4975/04/$30.00 doi:10.1016/j.athoracsur.2004.04.003
HAAN ET AL SURGERY FOR MR WITH LOW EF
821
Table 1. Patient Characteristics by Ejection Fraction Category Variable Age
Body surface area
Data Type Mean Median 25th 75th Mean Median 25th 75th
Cerebrovascular accident Chronic lung disease Diabetes Dialysis Pulmonary artery pressures Hypertension Intraaortic balloon pump/inotropic agents Infectious endocarditis Myocardial infarction New York Heart Association class IV Peripheral vascular disease Renal failure Status: Urgent Emergent Salvage
EF ⱕ 30% N ⫽ 727
EF ⬎ 30% N ⫽ 13,855
61.76 63.00 54.00 72.00 1.81 1.83 1.67 2.00 6.05 16.23 16.64 2.75 20.5 52.96 16.23 0.83 4.68 29.44 8.80 9.77
61.09 63.00 51.00 72.00 1.82 1.86 1.70 2.00 6.34 10.86 8.99 1.61 14.53 45.05 4.59 3.18 1.37 15.61 5.12 4.92
20.08 6.33 0.55
13.07 2.92 0.27
p value 0.2887
0.0166
0.7526 ⬍0.0001 ⬍0.0001 0.0191 ⬍0.0001 ⬍0.0001 ⬍0.0001 0.0003 ⬍0.0001 ⬍0.0001 ⬍0.0001 ⬍0.0001 ⬍0.0001
EF ⫽ ejection fraction.
2. If mitral valve surgery is reasonable, what impact does choice of procedure play? 3. How can one recognize the patient at particularly high risk, for whom the high perioperative mortality might recommend against valve replacement?
Patients and Methods Data Source The Society of Thoracic Surgeons (STS) National Cardiac Database was established in 1989 for the purpose of comparing outcomes following cardiac surgery [27]. Clinical patient data are entered at sites by using uniform definitions and certified software systems. Data from individual sites are harvested semiannually and sent to the STS Data Warehouse and Analysis Center at Duke Clinical Research Institute. A series of data quality checks are performed on the data, and then semiannual reports are sent back to the sites that compare their outcomes and patterns of resource utilization with those of their region and the nation. The accuracy and comparability of this voluntary database has been confirmed when compared with other mandatory and audited cardiac databases [28]. Ejection fraction is recorded as a measurement obtained by one or more of several methodologies: left ventriculogram, radionucleotide study, echocardiographic measurement, or an estimate from other calculations, based upon available clinical data. The evidence
of MR that is obtained from these preoperative studies is graded in the data collection tool as none, trivial, mild, moderate, or severe. Operative mortality is defined in the STS database as death within 30 days of surgery, whether in or out of the hospital, or death in the hospital regardless of length of stay. STS definitions of other preoperative risk factors, as well as complications, can be accessed on-line at (http://www.ctsnet.org/doc/2167).
Patient Population The study population consisted of patients undergoing an isolated mitral procedure (repair or replacement) between 1998 and 2001 for mild, moderate, or severe MR. Patients with prior cardiac surgery, a reported EF of less than 10%, and patients with missing data for age, gender, and EF were excluded from the analysis. Therefore, the target population of interest in this study consists of isolated primary mitral procedures for MR. The study population was not selected for nor analyzed by etiology of MR, but rather, the study looked at all patients coming to MV surgery and the outcomes thereof.
Statistical Analysis Because the ACC/AHA Guidelines use a 30% EF cutoff for management decision-making, we defined severe left ventricular dysfunction as an EF of 30% or less. Patient characteristics by these two EF groups (ⱕ 30 and ⬎ 30) were tabulated to better describe the differences between
CARDIOVASCULAR
Ann Thorac Surg 2004;78:820 –5
CARDIOVASCULAR
822
HAAN ET AL SURGERY FOR MR WITH LOW EF
Ann Thorac Surg 2004;78:820 –5
predictor variables that maximally distinguish mortality between the two branches.
Results Outcomes by Ejection Fraction
Fig 1. Observed percent mortality for mitral surgical procedures, by preoperative ejection fraction.
groups. We calculated operative mortality rates across six categories of EF to examine the relationship between EF and mortality. Next we calculated the unadjusted mortality and morbidity rates for the two EF groups and tested the differences with a 2 test. Major morbidities included stroke, need for reoperation, need for prolonged ventilation, deep sternal wound infection, new requirement for dialysis, and prolonged postoperative length of stay. A combined endpoint of the occurrence of any morbidity or mortality was also assessed. We then compared the observed morbidity outcomes for mitral surgery by EF above and below 30%. Multivariable logistic regression was used to determine the effect of severe left ventricular dysfunction on mortality. Variables from the STS Isolated Valve Surgery Risk Model [29] were included in the model for riskadjustment purposes. All variables from the valve model except EF were used to calculate a predicted risk for each patient. We then stratified patients into clinically similar and significant mortality risk groups (0% to 2.5%, 2.5% to 5.0%, 5% to 10%, 10% to 20%, and ⬎ 20%) and calculated their mortality rates within each of these risk groups. The low-EF patients were also compared for mortality outcomes by type of procedure to determine any differences between repair and replacement in LV dysfunction. To enable physicians to assign patients into relatively high-risk and low-risk subsets, a classification tree was developed by using the same set of variables used in the multivariate model. The classification tree identifies the
The STS National Cardiac Database contained 17,940 mitral procedures for MR that were performed between 1998 and 2001. After the appropriate exclusions, the resultant data set available for study contained 14,582 isolated, primary mitral procedures. As this study is of MR patients who underwent isolated MV procedures, it is of interest to look at MR grade distribution between the EF categories above or below 30%. Of the 13,855 patients with an EF of more than 30%, 668 were classified as having mild MR, 2456 had moderate MR, and 10,731 had severe MR. Of the 727 patients with an EF of 30% or less, 23 had mild MR, 121 had moderate MR, and 583 had severe MR. A comparison of patient characteristics by EF category, above or below 30%, is displayed in Table 1. Patients with an EF of 30% or less were more likely to have diabetes, renal failure, and hypertension. The low-EF patients also were more likely to be classified as New York Heart Association class IV, to have been given an intraaortic balloon pump or inotropic agents, or to be undergoing an emergent or salvage procedure. For analysis and display purposes, patients were divided into EF categories for easier comparison. The sample size for each EF group was 182 with an EF of 10% to 20%, 545 with an EF of 21% to 30%, 1384 with an EF of 31% to 40%, 3265 with an EF of 41% to 50%, and 9206 with EF of more than 50%. The percent mortality for each of these EF groups is shown in Figure 1. A distinct rise in mortality occurs in patients with a 30% EF, and it remains elevated at lower EFs. Unadjusted percent mortality for an EF of more than 30% is 3.14, compared with 5.35 for an EF of 30% or less (p ⫽ 0.0010). Acknowledging that patients undergoing isolated mitral valve surgery constitute a very heterogeneous group, we used the STS Valve Mortality Risk Model to account for differences in patient risk between the two groups. We found that a low EF was not a significant predictor of mortality after controlling for other patient factors (OR ⫽ 1.38; 95% CI, 0.96 to 1.99 and p ⫽ 0.09). To examine the effect more closely, we analyzed the patients by their predicted risk of mortality, grouped into
Table 2. Group Size for Mitral Procedure, by Ejection Fraction, by % Mortality Risk Ejection Fraction Group % Mortality Risk 0 –2.5 2.5–5.0 5.0 –10.0 10.0 –20.0 ⬎20.0 Total
10%–20%
21%–30%
31%– 40%
41%–50%
ⱖ50%
109 37 20 9 7 182
309 118 71 29 18 545
755 347 172 79 31 1384
2070 682 344 113 56 3265
6044 1815 876 330 141 9206
HAAN ET AL SURGERY FOR MR WITH LOW EF
Fig 2. Observed percent mortality following mitral valve procedure, by predicted risk group, by ejection fraction.
risk categories (Table 2). Using these groupings to describe more homogeneous subsets of MR patients, we looked at mortality by EF category in each risk group. Because of the very small cell sizes in the two highest risk categories, the individuals constituting risk groups of 10.0% to 20.0% and more than 20.0% were combined, and represented as a predicted risk of 10.0% or more. Figure 2 displays the relatively flat mortality curves, regardless of EF, in all risk groups except in those patients with a predicted mortality risk of more than 10.0%. The exception to this statement is the rise in mortality curve seen at 21% to 30% EF in patients with a predicted mortality risk of 5.0% to 10.0%. Even though mortality rises progressively for each risk group, mortality does not vary dramatically by EF until the predicted risk rises to more than 10.0%. Although the mortality for these higher risk patients is elevated for all EF categories, there is a rise with EF below 30% that continues steeply upward. The observed rates for each major morbidity, with associated p value, are described in Table 3. Patients with LV dysfunction tended to have higher rates of all major morbidities except renal failure, with the requirement for prolonged ventilation and prolonged length of stay reaching statistical significance.
Outcomes by Mitral Repair Versus Replacement We next sought to determine if outcomes differed appreciably by the type of procedure performed. Of the 14,582 Table 3. Unadjusted Rates for Morbidities After Mitral Surgery, by High and Low Ejection Fraction Outcome Stroke Reoperation Prolonged ventilator support Deep sternal wound infection Renal failure Combined morbidity and mortality Prolonged postop length of stay EF ⫽ ejection fraction.
EF ⱕ 30 EF ⬎ 30 p value 2.61 10.73 12.93 .069 3.32 24.07 11.42
1.67 8.66 7.28 0.28 5.94 16.51 8.17
.0578 .0546 ⬍.0001 .0516 .0002 ⬍.0001 .0020
823
Fig 3. Observed percent mortality for mitral surgical procedures, by type of procedure—replacement, repairs (all types), and repairs exclusive of annuloplasty-only repairs.
patients who had mitral valve surgery, 727 had an EF of 30% or less—339 had repairs and 388 had replacements. A comparison of mortality by repair versus replacement in patients with low EFs was performed. The percent operative mortality after repair was 1.40% to 1.53% for patients with an EF of 30% or less, while patients with mitral valve replacement had an increase in their mortality curve from about 4% up to 7.69% to 10.18% when the EF dropped below 30%. Acknowledging that mitral repairs are a heterogeneous group as to anatomy and etiology, we did an additional analysis to see if MR from dilated cardiomyopathy behaved differently with surgical procedure, thereby skewing the outcomes. Comparison of mortality for all mitral repairs with mortality for those repairs excluding annuloplasty-only procedures showed that both mortality curves are relatively flat at 2% or less (Fig 3). With relatively flat mortality curves for mitral repair, the rise in mortality rates seen in high-risk patients with low EFs, as displayed in Figure 2, is therefore, primarily attributable to the mitral replacement procedures.
What is the High-Risk Patient Profile? A subset of MR patients who undergo mitral valve surgery have a notably increased risk of mortality. In an effort to assist clinicians attempting to identify these low-EF patients at particularly high risk for operative mortality, a classification tree was used to delineate those risks noted in Table 1 that play the greatest role in mortality. As shown in Figure 4, age that exceeds 75 years, the presence of renal failure (especially on dialysis), and surgery done as an emergent or salvage procedure each contribute strongly to the mortality of low-EF patients undergoing mitral procedures. The CART model picks the age on the basis of the data that maximizes the difference in mortality rates (i.e., the difference in mortality rates for patients ⬍ 65 vs ⬎ 65 is not as big as the difference in mortality for patients ⬍ 75 vs ⬎ 75). Renal failure is characterized by a history of serum creatinine levels of 2.0 mg/dL or greater. Emergent
CARDIOVASCULAR
Ann Thorac Surg 2004;78:820 –5
CARDIOVASCULAR
824
HAAN ET AL SURGERY FOR MR WITH LOW EF
Ann Thorac Surg 2004;78:820 –5
patients with renal failure, or patients who require surgery as an emergent or salvage procedure. The impact of age on mitral valve surgery outcomes is consistent with the findings of Thourani and colleagues of Emory [30]. It is important to emphasize that this model and algorithm do not dictate treatment and management, but serve as one of many decision-making tools to assist the clinician.
Limitations
Fig 4. Operative mortality (%) for patients with severe left ventricular dysfunction undergoing mitral valve surgery. (CI ⫽ confidence interval.)
procedure is defined as one in which the patient has ongoing, refractory (difficult, complicated, or unmanageable), unrelenting cardiac compromise, and emergent salvage procedures are for those patients undergoing cardiopulmonary resuscitation en route to the operating room or before anesthesia induction. Eliminating these three risk factors lowers the risk of operative mortality to 1.7%, which makes the consideration of offering surgery for MR quite reasonable.
Comment The study shows that patients with decreased LV function (EF ⱕ 30%) and who have a preoperative risk of less than 10% can have mitral valve surgical procedures with acceptable perioperative mortality. Although this study does not mandate mitral repair where judged unsuitable or when the repair is not optimal, mitral valve repair can be offered to patients with MR and LV dysfunction without increased risk of mortality and morbidity, and where appropriate, is the first or optimal choice. In patients with low EFs and for whom adequate mitral repair is unsuitable or unavailable, mitral valve replacement can be done, albeit with somewhat increased risk of adverse outcomes. The STS Risk Model for valve surgery can be used to identify those patients at particularly higher risk and thus enable better patient selection in surgery for MR. However, for those clinicians who do not have ready access to the risk stratification capabilities of the STS database, or who seek guidance and greater ease in clinical decisionmaking for selecting patients appropriate for the consideration of mitral procedure, the classification tree we have provided can be put to use. Stated simply, mitral valve surgery in patients with severe LV dysfunction should be avoided, or at least very carefully weighed, in patients more than 75 years of age,
MR patients are a heterogeneous population in terms of etiology, pathophysiology, and procedure feasibility. This study does not stratify by MR etiology or timing of surgery in the patient’s disease process, nor can it speak to the variation among hospitals and surgeons with regard to skill, experience, and judgment in mitral valve surgery, especially involving repair. Our use of the risk stratification process is an attempt to diminish the amount of variability among patients, within and between groups, for better and more informative comparison of outcomes. More specific hemodynamic data, such as correlation of LV end-systolic volume index to EF categories and outcomes, are not available in this dataset. In addition, less specific or detailed interactions such as MR grade and EF categories on outcomes were also not performed. The additional stratification would have further divided groups that are already small enough to be of concern for statistical as well as clinical significance. Attention therefore remained focused on the influence of EF, acknowledging that other factors also influence outcomes in these patients. This study does, however, lay a foundation for the benefit of further study to help develop more detailed structure for clinical decision-making. The EF groups and risk groups do represent categorization of continuous variables, with lines connecting the data points for ease of interpretation, which carries the assumption that the continuous data act or perform along a continuum. Furthermore, there is a peak in the data of patients at 21% to 30% EF who are of moderate risk (5% to 10%), but this does not persist as a trend in rising mortality at lower EFs. Even in a database as large as the STS, a small numbers of patients with low EFs have undergone surgical procedures, especially in the higher risk groups. This implies the high degree of selection applied by both the referring cardiologists and the responsible surgeons. It is impossible to know how many MR patients were never referred for surgical consideration (perhaps based on the ACC/AHA Guidelines). It also does not account for those patients who were offered surgery but refused. Certainly as a surgical database, we are able to look only at the numerator. The high-risk patients with EFs of 30% or less represented only 5.5% of the patients in this study, certainly consistent with the currently biased selection for surgery, and making it difficult to detect clinically relevant differences between high-risk and low-risk patients. Those patients with low EFs who were operated on were deemed appropriate, in spite of the current guidelines, and it is from these patients that we seek to learn. Despite
the limits of this national database, it constitutes a larger number of patients than any single institution or region can assemble, and is therefore the best available dataset on which to build evidence-based recommendations and future prospective study. We conclude from the available data in this national database that mitral valve surgery does not present a prohibitive risk of operative mortality for appropriately selected patients. Mitral valve surgery should not be withheld from patients solely because of compromised LV function. This opinion stands in contrast to the current, available recommendations of the ACC/AHA Guidelines for valvular heart disease. It will be informative, going forward, to build on this information by further study using propensity matching and strata analysis for interaction of variables (e.g., EF and MR grade) and impact on outcomes. Additional study is also warranted to drill down on the etiologies of MR and the type of procedure performed, as relates to EF categories and to outcomes. Finally, an important question to clinicians and patients alike, but not addressed here, is that of the longterm outcome for mitral valve surgery in patients with MR and low EFs. It remains to be determined whether mitral valve surgery in ventricular dysfunction impacts long-term survival, congestive heart failure management, functional status, and quality of life for patients with MR.
References 1. Cooper HA, Gersh BJ. Treatment of chronic mitral regurgitation. Am Heart J 1998;135:925–36. 2. Acar J, Michel PL, Luxereau P, Vahanian A, Cormier B. Indications for surgery in mitral regurgitation. Eur Heart J 1991;12(SupplB):52–54. 3. Tribouilloy CM, Enriquez-Sarano M, Schaff HV, et al. Impact of preoperative symptoms on survival after surgical correction of organic mitral regurgitation—rationale for optimizing surgical indications. Circulation 1999;99:400 –5. 4. Bach DS, Bolling SF. Early improvement in congestive heart failure after correction of secondary mitral regurgitation in end-stage cardiomyopathy. Am Heart J 1995;129:1165–70. 5. Blondheim DS, Jacobs LE, Kotler MN, Costacurta GA, Parry WR. Dilated cardiomyopathy with mitral regurgitation: decreased survival despite a low frequency of left ventricular thrombus. Am Heart J 1991;122:763–71. 6. Phillips HR, Levine FH, Carter JE, et al. Mitral valve replacement for isolated mitral regurgitation: analysis of clinical course and late postoperative left ventricular ejection fraction. Am J Cardiol 1981;48:647–54. 7. Duarte IG, Murphy CO, Kosinski AS, et al. Late survival after valve operation in patients with left ventricular dysfunction. Ann Thorac Surg 1997;64:1089 –95. 8. Toeh KH, Ivanov J, Weisel RD. Determinants of survival and valve failure after mitral valve replacement. Ann Thorac Surg 1990;49:643–8. 9. Enriquez-Sarano M, Tajik AJ, Schaff HV, et al. Echocardiographic prediction of left ventricular function after correction of mitral regurgitation: results and clinical implications. J Am Coll Cardiol 1994;24:1536 –43. 10. Crawford MF, Souchek J, Oprian CA, et al. Determinants of survival and left ventricular performance after mitral valve
HAAN ET AL SURGERY FOR MR WITH LOW EF
11. 12.
13. 14. 15. 16. 17. 18. 19. 20.
21.
22.
23. 24. 25. 26. 27.
28. 29. 30.
825
replacement. Department of Veterans Affairs Cooperative Study on Valvular Heart Disease. Circulation 1990;81:1173– 81. Schlant RC. Timing of surgery for patients with nonischemic severe mitral regurgitation. Circulation 1999;99:338 –9. Bonow RO, Carabello B, DeLeon AC, et al. ACC/AHA 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 (Committee on Management of Patients with Valvular Heart Disease). J Am Coll Cardiol 1998;32:1486 –8. Bishay ES, McCarthy PM, Cosgrove DM, et al. Mitral valve surgery in patients with severe left ventricular dysfunction. Eur J Cardiothorac Surg 1000;17:213–21. Chen FY, Adams DH, Aranki SF, et al. Mitral valve repair in cardiomyopathy. Circulation 1998;98:II124 –7. Calafiore AM, Gallina S, DiMauro M, et al. Mitral valve procedure in dilated cardiomyopathy: repair or replacement? Ann Thorac Surg 2001;71:1146 –53. Bach DS, Bolling SF. Improvement following correction of secondary mitral regurgitation in end-stage cardiomyopathy with mitral annuloplasty. Am J Cardiol 1996;78:966 –9. Badhwar V, Bolling SF. Mitral valve surgery in the patient with left ventricular dysfunction. Semin Thorac Cardiovasc Surg 2002;14:133–6. Rothenburger M, Rukosujew A, Hammel D, et al. Mitral valve surgery in patients with poor left ventricular function. Thor Cardiovasc Surg 2002;50:351–4. Rocha ASC, Rocha NN, Soares RCV, et al. Improvement in left ventricular dysfunction after surgical correction of mitral regurgitation. Arq Bras Cardiol 2003;80:13–8. Tischler MD, Cooper KA, Rowen M, LeWinter MM. Mitral valve replacement versus mitral valve repair. A Doppler and quantitative stress echocardiographic study. Circulation 1994;89:132–7. Wencker D, Borer JS, Hochreiter C, et al. Preoperative predictors of late postoperative outcome among patients with nonischemic mitral regurgitation with ‘high risk’ descriptors and comparison with unoperated patients. Cardiology 2000;93:37–42. Enriquez-Sarano M, Tajik AJ, Schaff HV, Orszulak TA, Bailey KR, Frye RL. Echocardiographic prediction of survival after surgical correction of organic mitral regurgitation. Circulation 1994;90:830 –7. Gatti G, Cardu G, Pugliese P. Mitral valve surgery for mitral regurgitation in patients with advanced dilated cardiomyopathy. Ital Heart J 2003;4:29 –34. Bolling SF, Deeb GM, Brunsting LA, Bach DS. Early outcome of mitral valve reconstruction in patients with end-stage cardiomyopathy. J Thorac Cardiovasc Surg 1995;109:676 –83. Bolling SF, Pagani FD, Deeb GM, Bach DS. Intermediateterm outcome of mitral reconstruction in cardiomyopathy. J Thorac Cardiovasc Surg 1998;115:381–8. Enriquez-Sarano M, Schaff HV, Orszulak TA, Tajik AJ, Bailey KR, Frye RL. Valve repair improves the outcome of surgery for mitral regurgitation. Circulation 1995;91:1022–8. Boncheck LI, Wimer P. The Society of Thoracic Surgeons National Adult Cardiac Surgery Database and outcomes program participation manual. http://www.sts.org/file/ DatabaseManual2.pdf. 2001:1–15. Welke KF, Ferguson TB Jr, Coombs LP, et al. Validity of the Society of Thoracic Surgeons National Adult Cardiac Surgery Database. Ann Thorac Surg 2004;77(4):1137–9. Edwards FH, Peterson ED, Coombs LP, et al. Prediction of operative mortality after valve replacement surgery. J Am Coll Cardiol 2001;37:885–92. Thourani VH, Weintraub WS, Guyton RA, et al. Outcomes and long-term survival for patients undergoing mitral valve repair versus replacement. Circulation 2003;108:298 –304.
CARDIOVASCULAR
Ann Thorac Surg 2004;78:820 –5
Selecting Patients With Mitral Regurgitation and Left Ventricular Dysfunction for Isolated Mitral Valve Surgery Constance K. Haan, MD, Cristina I. Cabral, MD, Donald A. Conetta, MD, Laura P. Coombs, PhD, and Fred H. Edwards, MD Divisions of Cardiothoracic Surgery and Cardiology, University of Florida, Jacksonville, Florida, and The Outcomes Research and Assessment Group, Duke Clinical Research Institute, Durham, North Carolina
Background. American College of Cardiology/ American Heart Association (ACC/AHA) Guidelines state that patients with an ejection fraction (EF) of 30% or less should not undergo mitral valve replacement for mitral regurgitation (MR). We sought to establish, using a national cardiac surgery database, whether patients with left ventricular dysfunction may safely undergo mitral valve surgery for MR, and if so, which ones. Methods. We queried the Society of Thoracic Surgeons (STS) National Database to identify patients who had isolated mitral valve replacement or repair for MR between 1998 and 2001. Mortality and morbidity outcomes were compared by EF category (< 30% vs > 30%), and observed mortality compared by EF group, stratified by predicted risk for mortality. A classification and regression tree (CART) model was then used to determine which patient characteristics contributed most to designate the high-risk patient. Results. Of the 14,582 patients who had mitral valve
surgery, 727 had an EF of 30% or less and 13,855 had an EF of more than 30%. Observed mortality rates were higher for patients with an EF of 30% or less (5.4% vs 3.1%). However, for low-risk to medium-risk patients, mortality rates remained fairly constant across levels of EF. Mortality is notably increased in the high-risk patients (predicted risk > 10%). A classification tree identifies three key characteristics for high risk: age more than 75 years, renal failure, and emergent or salvage procedure. Conclusions. When the predicted mortality risk is less than 10%, EF has minimal impact on operative mortality for mitral regurgitation. In contrast to the ACC/AHA Guidelines, our data show that operative risk for mitral valve surgery is not prohibitive for most patients with ventricular dysfunction.
C
Accepted for publication April 1, 2004.
recommend, “if mitral valve repair appears likely, surgery should still be contemplated, provided ejection fraction is more than 30%.” [12] By contrast, some recent studies show that mitral valve surgery offers symptomatic improvement and survival benefit to MR patients with poor LV function [13–19]. Further, evidence indicates that mitral procedures in patients with MR and low EFs may be done with acceptable outcomes [13, 15, 20 –23]. It is known that operative outcomes associated with mitral valve repair are different from those associated with mitral valve replacement surgery [24 –26]. It is unclear how these outcome differences are influenced by patient variability and selection, or patient or surgeon bias for choice of operation, or both. Furthermore, most reports deal with small, and relatively select patient populations—with study patient populations ranging in size from 9 to 478 [2–10, 13–26]. We used a national database to study the inherently larger available number of patients to address the following questions:
Address reprint requests to Dr Haan, Division of Cardiothoracic Surgery, The Cardiovascular Center, UFHSC/J, 655 W Eighth St, Jacksonville, FL 32209; e-mail: [email protected].
1. Should mitral valve surgery be offered to patients with significant LV dysfunction?
ardiac surgery for mitral regurgitation (MR) has long been aimed toward intervening before significant deterioration of left ventricular (LV) function occurs [1]. This is intended to minimize perioperative morbidity and mortality and to optimize recovery and postoperative function [2– 4]. Mitral valve surgery in patients with MR and severe LV dysfunction has been associated with high operative mortality [5– 8]. Postoperative LV function and survival have been shown to be associated with preoperative LV function [6, 9, 10]. The firm recommendation has been made that MR patients with an ejection fraction (EF) of less than 30% should not be offered mitral valve surgery [11]. In fact, the ACC/AHA Guidelines for the Management of Patients with Valvular Heart Disease state that patients with an EF of less than 30% should not undergo mitral valve replacement for chronic MR, especially if the mitral valve apparatus [12] can not be preserved. Furthermore, the ACC/AHA Guidelines for Valve Surgery
© 2004 by The Society of Thoracic Surgeons Published by Elsevier Inc
(Ann Thorac Surg 2004;78:820 –5) © 2004 by The Society of Thoracic Surgeons
0003-4975/04/$30.00 doi:10.1016/j.athoracsur.2004.04.003
HAAN ET AL SURGERY FOR MR WITH LOW EF
821
Table 1. Patient Characteristics by Ejection Fraction Category Variable Age
Body surface area
Data Type Mean Median 25th 75th Mean Median 25th 75th
Cerebrovascular accident Chronic lung disease Diabetes Dialysis Pulmonary artery pressures Hypertension Intraaortic balloon pump/inotropic agents Infectious endocarditis Myocardial infarction New York Heart Association class IV Peripheral vascular disease Renal failure Status: Urgent Emergent Salvage
EF ⱕ 30% N ⫽ 727
EF ⬎ 30% N ⫽ 13,855
61.76 63.00 54.00 72.00 1.81 1.83 1.67 2.00 6.05 16.23 16.64 2.75 20.5 52.96 16.23 0.83 4.68 29.44 8.80 9.77
61.09 63.00 51.00 72.00 1.82 1.86 1.70 2.00 6.34 10.86 8.99 1.61 14.53 45.05 4.59 3.18 1.37 15.61 5.12 4.92
20.08 6.33 0.55
13.07 2.92 0.27
p value 0.2887
0.0166
0.7526 ⬍0.0001 ⬍0.0001 0.0191 ⬍0.0001 ⬍0.0001 ⬍0.0001 0.0003 ⬍0.0001 ⬍0.0001 ⬍0.0001 ⬍0.0001 ⬍0.0001
EF ⫽ ejection fraction.
2. If mitral valve surgery is reasonable, what impact does choice of procedure play? 3. How can one recognize the patient at particularly high risk, for whom the high perioperative mortality might recommend against valve replacement?
Patients and Methods Data Source The Society of Thoracic Surgeons (STS) National Cardiac Database was established in 1989 for the purpose of comparing outcomes following cardiac surgery [27]. Clinical patient data are entered at sites by using uniform definitions and certified software systems. Data from individual sites are harvested semiannually and sent to the STS Data Warehouse and Analysis Center at Duke Clinical Research Institute. A series of data quality checks are performed on the data, and then semiannual reports are sent back to the sites that compare their outcomes and patterns of resource utilization with those of their region and the nation. The accuracy and comparability of this voluntary database has been confirmed when compared with other mandatory and audited cardiac databases [28]. Ejection fraction is recorded as a measurement obtained by one or more of several methodologies: left ventriculogram, radionucleotide study, echocardiographic measurement, or an estimate from other calculations, based upon available clinical data. The evidence
of MR that is obtained from these preoperative studies is graded in the data collection tool as none, trivial, mild, moderate, or severe. Operative mortality is defined in the STS database as death within 30 days of surgery, whether in or out of the hospital, or death in the hospital regardless of length of stay. STS definitions of other preoperative risk factors, as well as complications, can be accessed on-line at (http://www.ctsnet.org/doc/2167).
Patient Population The study population consisted of patients undergoing an isolated mitral procedure (repair or replacement) between 1998 and 2001 for mild, moderate, or severe MR. Patients with prior cardiac surgery, a reported EF of less than 10%, and patients with missing data for age, gender, and EF were excluded from the analysis. Therefore, the target population of interest in this study consists of isolated primary mitral procedures for MR. The study population was not selected for nor analyzed by etiology of MR, but rather, the study looked at all patients coming to MV surgery and the outcomes thereof.
Statistical Analysis Because the ACC/AHA Guidelines use a 30% EF cutoff for management decision-making, we defined severe left ventricular dysfunction as an EF of 30% or less. Patient characteristics by these two EF groups (ⱕ 30 and ⬎ 30) were tabulated to better describe the differences between
CARDIOVASCULAR
Ann Thorac Surg 2004;78:820 –5
CARDIOVASCULAR
822
HAAN ET AL SURGERY FOR MR WITH LOW EF
Ann Thorac Surg 2004;78:820 –5
predictor variables that maximally distinguish mortality between the two branches.
Results Outcomes by Ejection Fraction
Fig 1. Observed percent mortality for mitral surgical procedures, by preoperative ejection fraction.
groups. We calculated operative mortality rates across six categories of EF to examine the relationship between EF and mortality. Next we calculated the unadjusted mortality and morbidity rates for the two EF groups and tested the differences with a 2 test. Major morbidities included stroke, need for reoperation, need for prolonged ventilation, deep sternal wound infection, new requirement for dialysis, and prolonged postoperative length of stay. A combined endpoint of the occurrence of any morbidity or mortality was also assessed. We then compared the observed morbidity outcomes for mitral surgery by EF above and below 30%. Multivariable logistic regression was used to determine the effect of severe left ventricular dysfunction on mortality. Variables from the STS Isolated Valve Surgery Risk Model [29] were included in the model for riskadjustment purposes. All variables from the valve model except EF were used to calculate a predicted risk for each patient. We then stratified patients into clinically similar and significant mortality risk groups (0% to 2.5%, 2.5% to 5.0%, 5% to 10%, 10% to 20%, and ⬎ 20%) and calculated their mortality rates within each of these risk groups. The low-EF patients were also compared for mortality outcomes by type of procedure to determine any differences between repair and replacement in LV dysfunction. To enable physicians to assign patients into relatively high-risk and low-risk subsets, a classification tree was developed by using the same set of variables used in the multivariate model. The classification tree identifies the
The STS National Cardiac Database contained 17,940 mitral procedures for MR that were performed between 1998 and 2001. After the appropriate exclusions, the resultant data set available for study contained 14,582 isolated, primary mitral procedures. As this study is of MR patients who underwent isolated MV procedures, it is of interest to look at MR grade distribution between the EF categories above or below 30%. Of the 13,855 patients with an EF of more than 30%, 668 were classified as having mild MR, 2456 had moderate MR, and 10,731 had severe MR. Of the 727 patients with an EF of 30% or less, 23 had mild MR, 121 had moderate MR, and 583 had severe MR. A comparison of patient characteristics by EF category, above or below 30%, is displayed in Table 1. Patients with an EF of 30% or less were more likely to have diabetes, renal failure, and hypertension. The low-EF patients also were more likely to be classified as New York Heart Association class IV, to have been given an intraaortic balloon pump or inotropic agents, or to be undergoing an emergent or salvage procedure. For analysis and display purposes, patients were divided into EF categories for easier comparison. The sample size for each EF group was 182 with an EF of 10% to 20%, 545 with an EF of 21% to 30%, 1384 with an EF of 31% to 40%, 3265 with an EF of 41% to 50%, and 9206 with EF of more than 50%. The percent mortality for each of these EF groups is shown in Figure 1. A distinct rise in mortality occurs in patients with a 30% EF, and it remains elevated at lower EFs. Unadjusted percent mortality for an EF of more than 30% is 3.14, compared with 5.35 for an EF of 30% or less (p ⫽ 0.0010). Acknowledging that patients undergoing isolated mitral valve surgery constitute a very heterogeneous group, we used the STS Valve Mortality Risk Model to account for differences in patient risk between the two groups. We found that a low EF was not a significant predictor of mortality after controlling for other patient factors (OR ⫽ 1.38; 95% CI, 0.96 to 1.99 and p ⫽ 0.09). To examine the effect more closely, we analyzed the patients by their predicted risk of mortality, grouped into
Table 2. Group Size for Mitral Procedure, by Ejection Fraction, by % Mortality Risk Ejection Fraction Group % Mortality Risk 0 –2.5 2.5–5.0 5.0 –10.0 10.0 –20.0 ⬎20.0 Total
10%–20%
21%–30%
31%– 40%
41%–50%
ⱖ50%
109 37 20 9 7 182
309 118 71 29 18 545
755 347 172 79 31 1384
2070 682 344 113 56 3265
6044 1815 876 330 141 9206
HAAN ET AL SURGERY FOR MR WITH LOW EF
Fig 2. Observed percent mortality following mitral valve procedure, by predicted risk group, by ejection fraction.
risk categories (Table 2). Using these groupings to describe more homogeneous subsets of MR patients, we looked at mortality by EF category in each risk group. Because of the very small cell sizes in the two highest risk categories, the individuals constituting risk groups of 10.0% to 20.0% and more than 20.0% were combined, and represented as a predicted risk of 10.0% or more. Figure 2 displays the relatively flat mortality curves, regardless of EF, in all risk groups except in those patients with a predicted mortality risk of more than 10.0%. The exception to this statement is the rise in mortality curve seen at 21% to 30% EF in patients with a predicted mortality risk of 5.0% to 10.0%. Even though mortality rises progressively for each risk group, mortality does not vary dramatically by EF until the predicted risk rises to more than 10.0%. Although the mortality for these higher risk patients is elevated for all EF categories, there is a rise with EF below 30% that continues steeply upward. The observed rates for each major morbidity, with associated p value, are described in Table 3. Patients with LV dysfunction tended to have higher rates of all major morbidities except renal failure, with the requirement for prolonged ventilation and prolonged length of stay reaching statistical significance.
Outcomes by Mitral Repair Versus Replacement We next sought to determine if outcomes differed appreciably by the type of procedure performed. Of the 14,582 Table 3. Unadjusted Rates for Morbidities After Mitral Surgery, by High and Low Ejection Fraction Outcome Stroke Reoperation Prolonged ventilator support Deep sternal wound infection Renal failure Combined morbidity and mortality Prolonged postop length of stay EF ⫽ ejection fraction.
EF ⱕ 30 EF ⬎ 30 p value 2.61 10.73 12.93 .069 3.32 24.07 11.42
1.67 8.66 7.28 0.28 5.94 16.51 8.17
.0578 .0546 ⬍.0001 .0516 .0002 ⬍.0001 .0020
823
Fig 3. Observed percent mortality for mitral surgical procedures, by type of procedure—replacement, repairs (all types), and repairs exclusive of annuloplasty-only repairs.
patients who had mitral valve surgery, 727 had an EF of 30% or less—339 had repairs and 388 had replacements. A comparison of mortality by repair versus replacement in patients with low EFs was performed. The percent operative mortality after repair was 1.40% to 1.53% for patients with an EF of 30% or less, while patients with mitral valve replacement had an increase in their mortality curve from about 4% up to 7.69% to 10.18% when the EF dropped below 30%. Acknowledging that mitral repairs are a heterogeneous group as to anatomy and etiology, we did an additional analysis to see if MR from dilated cardiomyopathy behaved differently with surgical procedure, thereby skewing the outcomes. Comparison of mortality for all mitral repairs with mortality for those repairs excluding annuloplasty-only procedures showed that both mortality curves are relatively flat at 2% or less (Fig 3). With relatively flat mortality curves for mitral repair, the rise in mortality rates seen in high-risk patients with low EFs, as displayed in Figure 2, is therefore, primarily attributable to the mitral replacement procedures.
What is the High-Risk Patient Profile? A subset of MR patients who undergo mitral valve surgery have a notably increased risk of mortality. In an effort to assist clinicians attempting to identify these low-EF patients at particularly high risk for operative mortality, a classification tree was used to delineate those risks noted in Table 1 that play the greatest role in mortality. As shown in Figure 4, age that exceeds 75 years, the presence of renal failure (especially on dialysis), and surgery done as an emergent or salvage procedure each contribute strongly to the mortality of low-EF patients undergoing mitral procedures. The CART model picks the age on the basis of the data that maximizes the difference in mortality rates (i.e., the difference in mortality rates for patients ⬍ 65 vs ⬎ 65 is not as big as the difference in mortality for patients ⬍ 75 vs ⬎ 75). Renal failure is characterized by a history of serum creatinine levels of 2.0 mg/dL or greater. Emergent
CARDIOVASCULAR
Ann Thorac Surg 2004;78:820 –5
CARDIOVASCULAR
824
HAAN ET AL SURGERY FOR MR WITH LOW EF
Ann Thorac Surg 2004;78:820 –5
patients with renal failure, or patients who require surgery as an emergent or salvage procedure. The impact of age on mitral valve surgery outcomes is consistent with the findings of Thourani and colleagues of Emory [30]. It is important to emphasize that this model and algorithm do not dictate treatment and management, but serve as one of many decision-making tools to assist the clinician.
Limitations
Fig 4. Operative mortality (%) for patients with severe left ventricular dysfunction undergoing mitral valve surgery. (CI ⫽ confidence interval.)
procedure is defined as one in which the patient has ongoing, refractory (difficult, complicated, or unmanageable), unrelenting cardiac compromise, and emergent salvage procedures are for those patients undergoing cardiopulmonary resuscitation en route to the operating room or before anesthesia induction. Eliminating these three risk factors lowers the risk of operative mortality to 1.7%, which makes the consideration of offering surgery for MR quite reasonable.
Comment The study shows that patients with decreased LV function (EF ⱕ 30%) and who have a preoperative risk of less than 10% can have mitral valve surgical procedures with acceptable perioperative mortality. Although this study does not mandate mitral repair where judged unsuitable or when the repair is not optimal, mitral valve repair can be offered to patients with MR and LV dysfunction without increased risk of mortality and morbidity, and where appropriate, is the first or optimal choice. In patients with low EFs and for whom adequate mitral repair is unsuitable or unavailable, mitral valve replacement can be done, albeit with somewhat increased risk of adverse outcomes. The STS Risk Model for valve surgery can be used to identify those patients at particularly higher risk and thus enable better patient selection in surgery for MR. However, for those clinicians who do not have ready access to the risk stratification capabilities of the STS database, or who seek guidance and greater ease in clinical decisionmaking for selecting patients appropriate for the consideration of mitral procedure, the classification tree we have provided can be put to use. Stated simply, mitral valve surgery in patients with severe LV dysfunction should be avoided, or at least very carefully weighed, in patients more than 75 years of age,
MR patients are a heterogeneous population in terms of etiology, pathophysiology, and procedure feasibility. This study does not stratify by MR etiology or timing of surgery in the patient’s disease process, nor can it speak to the variation among hospitals and surgeons with regard to skill, experience, and judgment in mitral valve surgery, especially involving repair. Our use of the risk stratification process is an attempt to diminish the amount of variability among patients, within and between groups, for better and more informative comparison of outcomes. More specific hemodynamic data, such as correlation of LV end-systolic volume index to EF categories and outcomes, are not available in this dataset. In addition, less specific or detailed interactions such as MR grade and EF categories on outcomes were also not performed. The additional stratification would have further divided groups that are already small enough to be of concern for statistical as well as clinical significance. Attention therefore remained focused on the influence of EF, acknowledging that other factors also influence outcomes in these patients. This study does, however, lay a foundation for the benefit of further study to help develop more detailed structure for clinical decision-making. The EF groups and risk groups do represent categorization of continuous variables, with lines connecting the data points for ease of interpretation, which carries the assumption that the continuous data act or perform along a continuum. Furthermore, there is a peak in the data of patients at 21% to 30% EF who are of moderate risk (5% to 10%), but this does not persist as a trend in rising mortality at lower EFs. Even in a database as large as the STS, a small numbers of patients with low EFs have undergone surgical procedures, especially in the higher risk groups. This implies the high degree of selection applied by both the referring cardiologists and the responsible surgeons. It is impossible to know how many MR patients were never referred for surgical consideration (perhaps based on the ACC/AHA Guidelines). It also does not account for those patients who were offered surgery but refused. Certainly as a surgical database, we are able to look only at the numerator. The high-risk patients with EFs of 30% or less represented only 5.5% of the patients in this study, certainly consistent with the currently biased selection for surgery, and making it difficult to detect clinically relevant differences between high-risk and low-risk patients. Those patients with low EFs who were operated on were deemed appropriate, in spite of the current guidelines, and it is from these patients that we seek to learn. Despite
the limits of this national database, it constitutes a larger number of patients than any single institution or region can assemble, and is therefore the best available dataset on which to build evidence-based recommendations and future prospective study. We conclude from the available data in this national database that mitral valve surgery does not present a prohibitive risk of operative mortality for appropriately selected patients. Mitral valve surgery should not be withheld from patients solely because of compromised LV function. This opinion stands in contrast to the current, available recommendations of the ACC/AHA Guidelines for valvular heart disease. It will be informative, going forward, to build on this information by further study using propensity matching and strata analysis for interaction of variables (e.g., EF and MR grade) and impact on outcomes. Additional study is also warranted to drill down on the etiologies of MR and the type of procedure performed, as relates to EF categories and to outcomes. Finally, an important question to clinicians and patients alike, but not addressed here, is that of the longterm outcome for mitral valve surgery in patients with MR and low EFs. It remains to be determined whether mitral valve surgery in ventricular dysfunction impacts long-term survival, congestive heart failure management, functional status, and quality of life for patients with MR.
References 1. Cooper HA, Gersh BJ. Treatment of chronic mitral regurgitation. Am Heart J 1998;135:925–36. 2. Acar J, Michel PL, Luxereau P, Vahanian A, Cormier B. Indications for surgery in mitral regurgitation. Eur Heart J 1991;12(SupplB):52–54. 3. Tribouilloy CM, Enriquez-Sarano M, Schaff HV, et al. Impact of preoperative symptoms on survival after surgical correction of organic mitral regurgitation—rationale for optimizing surgical indications. Circulation 1999;99:400 –5. 4. Bach DS, Bolling SF. Early improvement in congestive heart failure after correction of secondary mitral regurgitation in end-stage cardiomyopathy. Am Heart J 1995;129:1165–70. 5. Blondheim DS, Jacobs LE, Kotler MN, Costacurta GA, Parry WR. Dilated cardiomyopathy with mitral regurgitation: decreased survival despite a low frequency of left ventricular thrombus. Am Heart J 1991;122:763–71. 6. Phillips HR, Levine FH, Carter JE, et al. Mitral valve replacement for isolated mitral regurgitation: analysis of clinical course and late postoperative left ventricular ejection fraction. Am J Cardiol 1981;48:647–54. 7. Duarte IG, Murphy CO, Kosinski AS, et al. Late survival after valve operation in patients with left ventricular dysfunction. Ann Thorac Surg 1997;64:1089 –95. 8. Toeh KH, Ivanov J, Weisel RD. Determinants of survival and valve failure after mitral valve replacement. Ann Thorac Surg 1990;49:643–8. 9. Enriquez-Sarano M, Tajik AJ, Schaff HV, et al. Echocardiographic prediction of left ventricular function after correction of mitral regurgitation: results and clinical implications. J Am Coll Cardiol 1994;24:1536 –43. 10. Crawford MF, Souchek J, Oprian CA, et al. Determinants of survival and left ventricular performance after mitral valve
HAAN ET AL SURGERY FOR MR WITH LOW EF
11. 12.
13. 14. 15. 16. 17. 18. 19. 20.
21.
22.
23. 24. 25. 26. 27.
28. 29. 30.
825
replacement. Department of Veterans Affairs Cooperative Study on Valvular Heart Disease. Circulation 1990;81:1173– 81. Schlant RC. Timing of surgery for patients with nonischemic severe mitral regurgitation. Circulation 1999;99:338 –9. Bonow RO, Carabello B, DeLeon AC, et al. ACC/AHA 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 (Committee on Management of Patients with Valvular Heart Disease). J Am Coll Cardiol 1998;32:1486 –8. Bishay ES, McCarthy PM, Cosgrove DM, et al. Mitral valve surgery in patients with severe left ventricular dysfunction. Eur J Cardiothorac Surg 1000;17:213–21. Chen FY, Adams DH, Aranki SF, et al. Mitral valve repair in cardiomyopathy. Circulation 1998;98:II124 –7. Calafiore AM, Gallina S, DiMauro M, et al. Mitral valve procedure in dilated cardiomyopathy: repair or replacement? Ann Thorac Surg 2001;71:1146 –53. Bach DS, Bolling SF. Improvement following correction of secondary mitral regurgitation in end-stage cardiomyopathy with mitral annuloplasty. Am J Cardiol 1996;78:966 –9. Badhwar V, Bolling SF. Mitral valve surgery in the patient with left ventricular dysfunction. Semin Thorac Cardiovasc Surg 2002;14:133–6. Rothenburger M, Rukosujew A, Hammel D, et al. Mitral valve surgery in patients with poor left ventricular function. Thor Cardiovasc Surg 2002;50:351–4. Rocha ASC, Rocha NN, Soares RCV, et al. Improvement in left ventricular dysfunction after surgical correction of mitral regurgitation. Arq Bras Cardiol 2003;80:13–8. Tischler MD, Cooper KA, Rowen M, LeWinter MM. Mitral valve replacement versus mitral valve repair. A Doppler and quantitative stress echocardiographic study. Circulation 1994;89:132–7. Wencker D, Borer JS, Hochreiter C, et al. Preoperative predictors of late postoperative outcome among patients with nonischemic mitral regurgitation with ‘high risk’ descriptors and comparison with unoperated patients. Cardiology 2000;93:37–42. Enriquez-Sarano M, Tajik AJ, Schaff HV, Orszulak TA, Bailey KR, Frye RL. Echocardiographic prediction of survival after surgical correction of organic mitral regurgitation. Circulation 1994;90:830 –7. Gatti G, Cardu G, Pugliese P. Mitral valve surgery for mitral regurgitation in patients with advanced dilated cardiomyopathy. Ital Heart J 2003;4:29 –34. Bolling SF, Deeb GM, Brunsting LA, Bach DS. Early outcome of mitral valve reconstruction in patients with end-stage cardiomyopathy. J Thorac Cardiovasc Surg 1995;109:676 –83. Bolling SF, Pagani FD, Deeb GM, Bach DS. Intermediateterm outcome of mitral reconstruction in cardiomyopathy. J Thorac Cardiovasc Surg 1998;115:381–8. Enriquez-Sarano M, Schaff HV, Orszulak TA, Tajik AJ, Bailey KR, Frye RL. Valve repair improves the outcome of surgery for mitral regurgitation. Circulation 1995;91:1022–8. Boncheck LI, Wimer P. The Society of Thoracic Surgeons National Adult Cardiac Surgery Database and outcomes program participation manual. http://www.sts.org/file/ DatabaseManual2.pdf. 2001:1–15. Welke KF, Ferguson TB Jr, Coombs LP, et al. Validity of the Society of Thoracic Surgeons National Adult Cardiac Surgery Database. Ann Thorac Surg 2004;77(4):1137–9. Edwards FH, Peterson ED, Coombs LP, et al. Prediction of operative mortality after valve replacement surgery. J Am Coll Cardiol 2001;37:885–92. Thourani VH, Weintraub WS, Guyton RA, et al. Outcomes and long-term survival for patients undergoing mitral valve repair versus replacement. Circulation 2003;108:298 –304.
CARDIOVASCULAR
Ann Thorac Surg 2004;78:820 –5