|
|
||||||||
a Division of Cardiac Surgery, University of Ottawa, Ottawa, Ontario, Canada
b Division of Cardiology, University of Ottawa, Ottawa, Ontario, Canada
c Department of Epidemiology and Community Medicine, University of Ottawa, Ottawa, Ontario, Canada
Accepted for publication June 12, 2009.
* Address correspondence to Dr Ruel, University of Ottawa Heart Institute, 3403-40 Ruskin St, Ottawa, Ontario, K1Y 4W7, Canada (Email: mruel{at}ottawaheart.ca).
Presented at the Forty-fifth Annual Meeting of The Society of Thoracic Surgeons, San Francisco, CA, Jan 26–28, 2009.
| Dr Lam discloses that he has a financial relationship with Medtronic; Dr Ruel with Medtronic and St. Jude Medical.
|
| Abstract |
|---|
|
|
|---|
Methods: Between 1990 and 2005, 624 patients underwent MVR. Data included detailed serial echocardiographic tricuspid valve measurements, functional status, and survival data. Preoperative TR exceeded 2+ in 231: 125 received tricuspid repair and MVR, whereas 106 received MVR alone. Clinical and echocardiographic follow-up were complete (average, 6.8 ± 4.8 years). Parametric and semi-parametric tests were used to analyze outcomes.
Results: TR exceeding 2+ at operation was associated with a 53% increase in late death (p = 0.003). Tricuspid repair prevented echocardiographic progression of TR and improved congestive heart failure symptoms (both p < 0.01). Overall survival did not improve (p = 0.3). A trend to worsening TR in patients was noted with a larger tricuspid annulus diameter and without significant (
1+) TR preoperatively (odds ratio, 1.4 per cm increase in annulus diameter; p = 0.5), but this was not associated with worse functional or vital outcomes.
Conclusions: In patients undergoing MVR, tricuspid repair is indicated when TR exceeds 2+ to alleviate heart failure symptoms, but without significantly improving survival in this population. TR of 2+ or less may not require repair. Echocardiographic tricuspid annular dimensions alone, in the absence of significant (
1+) TR preoperatively, should not dictate the performance of tricuspid repair.
Functional tricuspid regurgitation (TR) is common in patients undergoing operations for mitral valve disease, with a reported prevalence between 25% and 30% [1–3]. Functional TR has a number of causes related to dilation of the tricuspid annulus and tethering of the tricuspid valve leaflets secondary to right ventricular (RV) dysfunction [4, 5]. Historically, concomitant tricuspid valve repair was less commonly performed due to the observation that functional TR often improves after mitral operations [6]. Contemporary evidence, however, suggests that tricuspid repair should be considered more often given that uncorrected functional TR is associated with poor survival and functional status [1–3, 7, 8].
Although functional TR is associated with negative outcomes, it remains unclear whether surgical correction translates into any improvement in functional status or survival. In addition, evidence is scarce in advising a minimum threshold of TR severity for which the clinical benefit of tricuspid valve repair may be observed [9].
The objective of this study was to determine in patients undergoing mitral valve replacement (MVR), the influence of preoperative TR and concomitant tricuspid valve repair on survival, persistence or recurrence of congestive heart failure (CHF), and postoperative echocardiographic TR progression. To this end, we studied 624 MVR patients with varying degrees of concomitant preoperative TR and who underwent tricuspid valve repair or had no repair.
| Material and Methods |
|---|
|
|
|---|
Patient Population
Between 1990 and 2005, 624 patients underwent MVR with commercially available prostheses. The preoperative TR grade exceeded 2+ in 231 patients. Of these, 125 underwent tricuspid valve repair, which consisted of the DeVega annuloplasty technique in 28 (22%), ring annuloplasty alone in 85 (68%), or a combination of ring annuloplasty and valvuloplasty in 12 (10%; Table 1,
Fig 1).
|
|
Operative Characteristics
All patients underwent MVR through median sternotomy, with the use of cardiopulmonary bypass, mild systemic hypothermia, and cold blood or crystalloid cardioplegic arrest. MVR was performed though a left atrial incision; the posterior leaflet of the mitral valve was preserved whenever possible, and sizing and orientation of the prosthesis were performed in accordance with manufacturers' guidelines. Braided 2-0 sutures with pledgets were used for mitral prosthesis insertion. Concomitant coronary artery bypass grafting (CABG) was performed on coronary vessels 1.5 mm or larger in diameter that displayed a luminal stenosis exceeding 50% on preoperative angiography.
Selection of the valve replacement prosthesis was performed according to each patient's informed preferences. The implanted mitral prostheses were CarboMedics (Sorin Biomedica, Via Crescentino, Italy), Carpentier Edwards (CE) Perimount (Edwards Lifesciences, Irvine, CA), Medtronic Hancock II, Medtronic Hall (Medtronic, Minneapolis, MN); On-X (MCRI, Austin, TX); and St. Jude Medical (St. Jude Medical, St. Paul, MN; Table 2).
|
Follow-Up
Patients were assessed annually in a dedicated valve clinic. Clinic visits consisted of a history and physical examination, electrocardiogram, chest roentgenogram, complete blood count, serum chemistries, and international normalized ratio, when applicable. Prosthesis-related complications were recorded according to the "Guidelines for Reporting Mortality and Morbidity After Cardiac Valve Interventions" [10]. Mean follow-up was 6.8 ± 4.8 years, and was complete for all patients.
Echocardiographic Assessment
All patients underwent a M-mode 2-dimensional, and Doppler transthoracic or transesophageal echocardiogram, or both, to measure preoperative cardiac dimensions, transvalvular pressure gradients, valve regurgitation severity, and RV systolic pressure (RVSP) as recommended by the American Society of Echocardiography (ASE) [11–13]. The tricuspid annulus diameter was measured as the maximum distance between the septal and lateral leaflet-annular insertion points in the apical 4-chamber view during diastole according to ASE guidelines [11–13]. TR severity was quantified as none to trivial (0), mild (1+), mild to moderate (2+), moderate to severe (3+), and severe (4+) using ASE recommended criteria [11].
RV function was quantified by the degree of dysfunction as none, mild, moderate, or severe. RVSP was measured using the equation: RVSP = 4 VTR 2+ RAP, where VTR is the maximum TR velocity obtained by continuous-wave Doppler and RAP is an estimate of right atrial pressure [13].
Postoperatively, all patients underwent regular follow-up transthoracic echocardiograms using these methods at their first annual visit, and subsequently according to valve clinic guidelines or as clinically indicated.
Statistical Analyses
Data were imported and analyzed in Stata 10.1 software (College Station, TX). Patients were compared with respect to (1) baseline patient, operative, and echocardiographic variables; (2) presence of CHF and all-cause mortality; and (3) degree and predictors of TR progression, based on preoperative TR severity and presence of tricuspid valve repair.
Continuous data are described as mean ± standard deviation. A
2 or Fisher exact test was used for categoric variables. A one-way analysis of variance with Bonferroni correction was used for continuous variables. Statistical significance was set at p < 0.05.
A log-rank test was used to determine equality of survival and CHF predictors. CHF was defined as New York Heart Association class III or IV symptoms present for 4 consecutive weeks at any point postoperatively or where the primary cause of death was CHF. For multivariable models, the proportional hazard assumption was verified with generalized Cox-Snell residuals. If the assumption was met, Cox proportional hazards models were developed by incorporating into each model (1) variables with p
0.1 on log-rank testing, (2) patient and operative characteristics that may have differed according to the degree of preoperative TR, (3) previously described predictors of CHF or death for patients undergoing MVR [14]. All covariates were used simultaneously to better account for confounding. Predictors of TR progression were determined by use of a logistic regression model. In patients with multiple postoperative echocardiographic measurements, the most severe TR recorded was selected as the response variable.
| Results |
|---|
|
|
|---|
Effect of Preoperative TR Severity on Survival
Preoperative TR severity affects survival, but for a given degree of preoperative TR, concomitant tricuspid valve repair does not. The overall 30-day mortality of 3% (18 of 624) was not significantly different between patient groups (Table 2). Figure 2
describes the actuarial survival of patients by preoperative TR severity and the presence of concomitant tricuspid valve repair. Patients with preoperative TR exceeding 2+ had worse survival relative to patients with TR of 2+ or less who did not receive tricuspid repair (hazard ratio [HR], 1.5; 95% confidence interval [CI] 1.1 to 2.0; p = 0.003). Notably, patients with preoperative TR exceeding 2+ who received tricuspid repair had comparable survival to patients without tricuspid repair, with 5- and 10-year actuarial survivals of 77.9% ± 3.9% and 62.5% ± 5.6% vs 80.2% ± 4.2% and 65.9% ± 5.5%, respectively (p = 0.3).
|
|
|
|
|
|
| Comment |
|---|
|
|
|---|
Patients with preoperative TR exceeding 2+ were older, more likely women, and more often in atrial fibrillation preoperatively compared with patients with less severe TR. Other authors have also observed this demographic trend [15]. Patients with preoperative TR exceeding 2+ more often had echocardiographic evidence of chronic left-heart overload such as left atrial and tricuspid annulus enlargement, as well as higher right ventricular systolic pressures. Among patients with preoperative TR exceeding 2+, those who received concomitant tricuspid repair were more likely to have had 4+ MR preoperatively.
The first important finding of this study was the observation that concomitant tricuspid valve repair did not significantly affect early or late survival. No difference existed in the early mortality rate between patient groups, with an overall early mortality of 3% that is comparable to other studies [7, 16, 17]. Furthermore, the 5- and 10-year actuarial survival of patients in our series is consistent with previously published data [7, 16, 17] and showed no adverse vital effect of conservatively managing TR at the time of MVR. Of note is that these observations may in part reflect selection bias, considering that patients with a preoperative TR exceeding 2+ who received concomitant tricuspid repair were less likely to have undergone MVR for ischemic mitral regurgitation.
Dreyfus and colleagues [17] have described the long-term outcomes of patients undergoing tricuspid repair at time of MVR. In their study, all patients undergoing MVR with an intraoperatively measured tricuspid annulus exceeding 70 mm, as measured from the anteroseptal commissure to the posteroseptal commissure, received tricuspid valve repair. After a mean follow-up of 4.8 years, no difference in survival was observed between patients receiving MVR alone vs MVR and concomitant tricuspid repair [17].
Other predictors of survival in this study included age, concomitant CABG, preoperative NYHA class, and left ventricular function, all of which have previously been shown to affect late survival in patients undergoing MVR [7, 14, 16].
Despite the lack of a significant effect on survival, tricuspid valve repair prevented persistent or recurrent CHF in patients undergoing MVR, which constitutes an important finding documenting the benefits of tricuspid valve repair on functional status. Of note is that Dreyfus and colleagues [18] described an improvement in NYHA class in patients receiving tricuspid repair and MVR at midterm follow-up. Their study, however, did not describe the outcomes of patients with large tricuspid annulus measurements who did not receive tricuspid valve repair, thereby preventing an appropriate counterfactual comparison between patients receiving and those not receiving tricuspid repair. Other predictors of persistent or late CHF in our study included age and left ventricular function, which are consistent with previous reports [14].
To further understand the effect of concomitant tricuspid repair on late outcomes, its influence on postoperative TR was also analyzed. Not surprisingly, tricuspid valve repair prevents TR progression over time; in fact, performance of tricuspid repair was the most important independent predictor of TR progression. Left atrial enlargement and preoperative atrial fibrillation have been associated with persistent TR in patients after left-heart valve surgery; however, these studies did not include concomitant tricuspid valve repair [7, 19, 20]. Notably, our data revealed that tricuspid annulus diameter did not affect TR progression when TR did not exist preoperatively, which contrasts with some previously published data [17]. On the other hand, our study found that 3+ or 4+ TR was more likely to develop in patients with the largest quartile of tricuspid annulus size when the preoperative TR exceeded 2+.
Limitations
This study is a retrospective analysis of prospective cohort data, and therefore unknown confounders may have influenced outcomes beyond that controlled in the multivariate analysis. Also, the study interval spans 15 years, and several tricuspid repair techniques in addition to mitral valve prostheses were used during this time period. Although detailed serial echocardiographic measurements were obtained, detailed tricuspid valve measurements such as the RV sphericity index or tricuspid leaflet tethering were not measured. These have been hypothesized to influence residual TR after mitral operations [21] and might provide additional interesting insight. Overall, tricuspid valve repair was not associated with an improvement in vital outcomes in this study. It remains unclear whether a difference would have been observed with a larger patient cohort.
Recommendations
In patients undergoing MVR, tricuspid repair is indicated when TR exceeds 2+ to alleviate CHF symptoms, but without significantly improving survival. TR at 2+ or less may not require repair. Furthermore, echocardiographic tricuspid annular dimensions alone, in the absence of significant (
1+) TR preoperatively, should not dictate the performance of tricuspid repair.
| Discussion |
|---|
|
|
|---|
DR CHAN: That is an excellent question, which addresses likely the most controversial part of this presentation. In our study, the tricuspid annulus diameter was measured as the distance between the septal and the lateral leaflets via the apical 4-chamber view. The normal reference value for the tricuspid annulus diameter in this view is approximately 2.2 ± 0.3 cm. Our measurements for each of the three study groups were larger than normal likely as a consequence of the underlying mitral valve disease.
The best way that I can answer your question is to relate the influence of tricuspid annular size and preoperative TR [tricuspid regurgitation] severity on the development of 3 or 4+ TR. Shown here is a table describing the number of patients developing postoperative 3 or 4+ TR among patients with preoperative TR equal to 0, and who did not receive tricuspid valve repair. Instead of using a strict cutoff, we divided our patient population into quartiles based on the tricuspid annulus diameter measurement. A tricuspid annulus measurement of 2.6 cm or less was in the lowest quartile, whereas a tricuspid annulus measurement greater than 3.4 was in the highest quartile. The proportion of patients that developed 3 or 4+ TR was not significantly different based on tricuspid annulus measurements. And the same observation was also made in regards to patients with preoperative TR severity of 1+ and also 2+. However, in patients with TR equal to 3+ preoperatively, patients with larger tricuspid annulus diameters were more likely to develop postoperative TR of 3 or 4+.
DR DION: Thank you very much.
DR DAVID H. ADAMS (New York, NY): I enjoyed your talk as well. Can you tell me about the etiology of mitral valve disease? Were these patients a mixed bag and did you consider that in your analysis?
DR CHAN: That is an excellent question. We divided our patient population based on the type of echocardiographically determined mitral valve lesion. Specifically, patients with mitral stenosis, concomitant or otherwise, were not included in this analysis due to the possibility of rheumatic heart disease.
DR ADAMS: Well that is important, because you need to control for ischemic mitral regurgitation vs degenerative. Were there any valve repairs in this series?
DR CHAN: This cohort refers only to patients that received mitral valve replacement and does not include patients with mitral valve repair. To answer your first question, within our multivariate modeling we did include concomitant coronary revascularization as a covariate.
DR ADAMS: That is just a surrogate. We really need to know in the final analysis how many patients had degenerative disease versus restricted leaflet motion from remodeling.
DR CHAN: Unfortunately, the specifics of leaflet function are not clearly documented in all of our echocardiographic reports.
DR ADAMS: Thank you.
| Acknowledgments |
|---|
|
|
|---|
| References |
|---|
|
|
|---|
This article has been cited by other articles:
![]() |
M. Taramasso, H. Vanermen, F. Maisano, A. Guidotti, G. La Canna, and O. Alfieri The Growing Clinical Importance of Secondary Tricuspid Regurgitation J. Am. Coll. Cardiol., February 21, 2012; 59(8): 703 - 710. [Abstract] [Full Text] [PDF] |
||||
![]() |
J. B. Kim, D. G. Yoo, G. S. Kim, H. Song, S.-H. Jung, S. J. Choo, C. H. Chung, and J. W. Lee Mild-to-moderate functional tricuspid regurgitation in patients undergoing valve replacement for rheumatic mitral disease: the influence of tricuspid valve repair on clinical and echocardiographic outcomes Heart, January 1, 2012; 98(1): 24 - 30. [Abstract] [Full Text] [PDF] |
||||
![]() |
J. M. Stulak, R. M. Suri, J. A. Dearani, T. M. Sundt III, and H. V. Schaff When Should Prophylactic Maze Procedure Be Considered in Patients Undergoing Mitral Valve Surgery? Ann. Thorac. Surg., May 1, 2010; 89(5): 1395 - 1401. [Abstract] [Full Text] [PDF] |
||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| 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 |