Restoration of Sinus Rhythm by the Maze Procedure Halts Progression of Tricuspid Regurgitation After Mitral Surgery

Transcription

1 Restoration of Sinus Rhythm by the Maze Procedure Halts Progression of Tricuspid Regurgitation After Mitral Surgery John M. Stulak, MD,* Hartzell V. Schaff, MD, Joseph A. Dearani, MD, Thomas A. Orszulak, MD, Richard C. Daly, MD, and Thoralf M. Sundt III, MD Division of Cardiovascular Surgery, Mayo Clinic and Foundation, Rochester, Minnesota Background. Tricuspid regurgitation (TR) is common in patients undergoing mitral valve surgery, and atrial fibrillation (AF) can cause progression of TR. This study examined the hypothesis that correction of AF with the Maze procedure can prevent the late progression of TR after mitral valve surgery. Methods. We compared severity of TR in 33 patients (January 1, 1993, to January 1, 2003) who underwent concomitant Maze procedure plus mitral valve surgery and converted to normal sinus rhythm postoperatively with case-matched control patients who underwent mitral valve surgery alone and remained in AF postoperative. Matched variables were age, sex, diabetes mellitus, left ventricular ejection fraction, and hypertension. Preoperative TR grade was similar between groups (Maze versus no-maze , p 0.67). Patients with permanent transvenous pacemakers, organic tricuspid valve disease, and prior tricuspid valve surgery were excluded from this comparison. Results. Before hospital dismissal, average TR grade improved to in both groups; TR improved in 42% of patients in the Maze group and 36% of patients with preoperative AF and no Maze. At last follow-up, average TR grade remained stable at in the Maze group (p versus preoperative) with TR progression in only 9% of patients (3 of 33). In contrast, TR grade worsened to in the no-maze group (p 0.04 versus preoperative, p < versus postoperative, p < versus groups), and TR worsened in 45% of patients (15 of 33). In a multivariable model, performance of a Maze procedure was protective against the progression of TR. Conclusions. Continued AF after mitral valve surgery can predispose a patient to progression of TR, and this progression is prevented in patients having successful concomitant Maze procedure. (Ann Thorac Surg 2008;86:40 5) 2008 by The Society of Thoracic Surgeons Tricuspid regurgitation (TR) may be present in 10% to 50% of patients with significant mitral valve regurgitation or stenosis, or both, as well as other left-sided valve lesions [1 3]. In most patients, concomitant TR is a functional rather than an organic phenomenon, and is commonly associated with elevated pulmonary arterial pressure and right ventricular dilatation [4]. Although TR may decrease after correction of left-sided cardiac pathology [5], in some patients, TR persists or worsens [6]. The course of TR after mitral valve surgery is difficult to predict. It is known, however, that the persistence or late occurrence of TR after correction of mitral valve surgery contributes to significant morbidity and mortality [7, 8], and reoperation to correct TR in this setting carries a high operative mortality and poor functional results [7, 9]. Accepted for publication March 3, *Recipient of the 2007 Southern Thoracic Surgical Association President s Award. Presented at the Fifty-fourth Annual Meeting of the Southern Thoracic Surgical Association, Bonita Springs, FL, Nov 7 10, Address correspondence to Dr Schaff, Division of Cardiovascular Surgery, Mayo Clinic College of Medicine, 200 First St SW, Rochester, MN 55905; schaff@mayo.edu. Thus, it is important to understand the mechanism late TR and to develop strategies to prevent its progression. Atrial fibrillation (AF) is common in patients with left-sided valvular diseases, and the arrhythmia often persists despite successful correction of mitral or aortic pathology, leading to poor outcome [10, 11]. In addition, AF has been identified as a significant predictor for the late development of TR after mitral valve surgery [6]. The Cox Maze procedure has proven safe and extremely effective for the surgical treatment of atrial fibrillation [12 14], and is most often used in patients undergoing mitral valve operations to enhance the probability restoring sinus rhythm postoperatively [15, 16]. We conducted this study to investigate the hypothesis that the restoration of sinus rhythm with the Maze procedure at the time of mitral valve surgery may prevent progression of TR and its associated clinical sequelae. Patients and Methods The Mayo Foundation Institutional Review Board approved this study, and all patients or their families gave written informed consent. After IRB approval, we reviewed medical records for patient demographics, past 2008 by The Society of Thoracic Surgeons /08/$34.00 Published by Elsevier Inc doi: /j.athoracsur

2 Ann Thorac Surg STULAK ET AL 2008;86:40 5 MAZE PROCEDURE AF CORRECTION HALTS TR PROGRESS medical and surgical history, operative procedure, preoperative and postoperative cardiac rhythm, preoperative and postoperative echocardiographic tricuspid and mitral valve analysis, early and late morbidity, and survival. To investigate the effect of atrial fibrillation after correction of mitral valve pathology on progression of tricuspid regurgitation, we identified 33 patients with preoperative AF who underwent combined Maze procedure and mitral valve operation who converted to sinus rhythm postoperatively and remained in sinus rhythm through last follow-up. These patients were compared with case-matched controls in AF preoperatively who underwent mitral valve surgery alone during the same time period and remained in AF postoperatively and through last follow-up. Patients with permanent transvenous pacemakers, organic tricuspid valve disease, and prior tricuspid valve surgery were excluded from this study. Matched variables were age (Maze, 63 years; no-maze, 65 years; p 0.54), sex (Maze, 24 males; no-maze, 19 males; p 0.2), diabetes mellitus (Maze, 1 patient; no- Maze, 2 patients; p 0.56), left ventricular (LV) ejection fraction (Maze, 58%; no-maze, 57%; p 0.85), and hypertension (Maze, 11 patients; no-maze, 12 patients; p 1). We chose these variables to arrive at a patient population that was clinically similar in terms of cardiac function and comorbidities that are known to influence AF. We did not match according to duration of AF, a known factor influencing the success of the Maze procedure, because this information was either unclear or unknown in several of our patients, thus making it an unreliable variable on which to match patient populations. In addition, we chose not to use left atrial size as a matching criteria because the majority of patients in both groups already had enlarged atria and were, thus, similar already in that regard. Unmatched clinical variables were generally similar in the two groups, aside from a slightly worse preoperative New York Heart Association function class in the no- Maze group (2.4 versus 1.8, p 0.001). Echocardiographic data were similar between study groups, including left atrial dimension (Maze, 55.9 mm; no-maze, 62.5 mm; p 0.052), LV end-diastolic dimension (Maze, 57.1 mm; no- Maze, 55.9 mm; p 0.62), systolic pulmonary artery pressure (45 mm Hg in both groups, p 0.88), preoperative MR grade (Maze, 3.7; no-maze, 3.5; p 0.67), and preoperative TR grade (Maze, 2.2; no-maze, 2.3; p 0.67). Right ventricular size and function were also similar between study groups. Two-dimensional and Doppler echocardiographic evaluation was performed before hospital dismissal and at late follow-up for measurement of cardiac function and dimensions, estimation of pulmonary artery pressures, and assessment of degree of valvular regurgitation. Severity of regurgitation was classified as 0 (none), 1 (trivial), 2 (mild), 3 (moderate), 3.5 (moderate-severe), and 4 (severe). Demographic and other patient-related data were obtained from Mayo Clinic medical records. Follow-up information was obtained from subsequent Table 1. Concomitant Procedures Performed for Both the Maze and No Maze Study Groups Procedure clinic visits, written correspondence from local physicians, and mailed questionnaires to patients or families. Descriptive statistics for categorical variables are reported as frequency and percentage while continuous variables are reported as mean (standard deviation) or median (range) as appropriate. Categorical variables were compared between cases and controls using a 2 test, and continuous variables were compared using a two sample t test or Wilcoxon rank-sum test where appropriate. While clinical experience demonstrates that the TR grades are not always equally spaced, and is subjective in judgment, we converted the grade into scale of 0, 1, 2, 3, 4, and thus treat it as a continuous variable. The within-group changes of TR were tested using paired t test or Wilcoxon signed-rank test, while the betweengroup comparison of changes were tested using twosample t test or Wilcoxon rank-sum test where appropriate. We reviewed the charts, and if a patient s TR grade increased, it was considered an event of TR progression; if TR grade decreased, it was similarly treated as TR regression. Accordingly, logistic models were used to identify the predictors for the TR progression. The multivariable model considered univariately significant variables (p 0.05) with model selection using the stepwise method (backward and forward methods resulted in the same model). All statistical tests were two-sided with the alpha level set at 0.05 for statistical significance. Results Maze (10 Patients) No Maze (14 Patients) Aortic valve surgery 2 6 Coronary bypass grafting 6 7 Atrial septal defect closure 1 2 Elastoma resection 1 0 Septal myectomy 0 1 Left ventricular aneurysm 0 1 resection Mitral annular decalcification 0 1 Total All patients in both groups had preoperative atrial fibrillation (longstanding persistent in 30 patient in each group) and underwent surgery on the mitral valve. Patients in the Maze group underwent concomitant biatrial cut-and-sew Maze procedure along with other cardiac surgical procedures in 10 patients (30%), and 14 patients (42%) underwent concomitant cardiac procedures in the no-maze group (Table 1). Surgery on the mitral valve included mitral repair in 31 of 33 (93%) Maze patients and in 14 of 33 (42%) no-maze patients. When the mitral valve was replaced, a mechanical prosthesis was used in 1 of 2 patients (50%) in the Maze group, and in 14 of 19 (74%) in the no-maze group. The surgical approach to the mitral valve was made by the operating surgeon after assess- 41

3 42 STULAK ET AL Ann Thorac Surg MAZE PROCEDURE AF CORRECTION HALTS TR PROGRESS 2008;86:40 5 Fig 1. Left atrium size. Change in left atrial dimension for both the Maze (solid circles) and no-maze (solid boxes) groups during the preoperative (Preop), postoperative (Postop), and follow-up (FU) periods. ment of valve anatomy and pathology present. Aortic cross-clamp and cardiopulmonary bypass times were longer in patients having concomitant Maze procedures (cross-clamp 60 minutes versus 52 minutes, p 0.18; cardiopulmonary bypass 103 minutes versus 73 minutes, p 0.01). Clinical follow-up was obtained in 29 of 33 patients (88%) in the Maze group and in 32 of 33 (97%) in the no-maze group. Median follow-up in the Maze group was 4.9 years (range, 1 month to 10.2 years), and 8.3 years (range, 1 month to 14.3 years) in the no-maze group (p 0.5). There were no late deaths in the Maze group and 2 late deaths in the no-maze group (p 0.16). Echocardiographic follow-up was available in all patients for whom clinical follow-up was obtained, and was 3.8 years (range, 1 month to 10.3 year) in the Maze group, and 6.6 years (range, 1 month to 13.6 years) in the no-maze group (p 0.5). Significant improvement in mitral regurgitation was seen in both the Maze group (3.7 preoperative to 1.4 late, p 0.01) and the no-maze group (3.5 preoperative to 1.1 late, p 0.01; p 0.17 between groups at late follow-up). Early postoperatively, left atrial size decreased to 49.2 mm (versus 57.1 mm preoperative, p 0.41) in the Maze group and 51.8 mm (versus 55.9 mm preoperative, p Fig 3. Progression of tricuspid regurgitation. Percentage of patients with progression of tricuspid regurgitation for both the Maze (hatched bars) and no-maze (solid bars) groups for the preoperative (Preop) to postoperative (Postop), postoperative to follow-up (FU), and preoperative to follow-up periods ) in the no-maze group (p not significant between groups; Fig 1). At late follow-up, left atrial size remained stable in the Maze group, but increased in the no-maze group (p 0.05 versus groups; Fig 1). At last echocardiographic follow-up, systolic pulmonary artery pressure significantly decreased to 39 mm Hg (versus 45 mm Hg preoperative, p 0.04) in the Maze group, while remaining at 45 mm Hg (p 0.99) in the no-maze group. Among patients having a concomitant Maze operation, right ventricular size decreased in 2 patients, increased in 5, and remained stable in 22; in the no-maze patients, right ventricular size decreased in 1, increased in 12, and remained stable in 17 (p between groups). In the Maze patients, right ventricular function improved in 3 patients, worsened in 6, and remained stable in 20; whereas no patients in the no-maze group improved, 8 worsened, and 22 remained stable (p 0.6 between groups). Before hospital dismissal, TR grade improved to 1.9 in both groups (p 0.77). At last follow-up, TR grade remained stable at 1.9 in the Maze group (p versus preoperative) and worsened to 2.7 in the no-maze group (p versus groups; Fig 2). Tricuspid valve regurgitation progressed in 9% (3 of 33) in the Maze group compared with 45% (15 of 33) in the no-maze group (Fig 3). Because of small differences in hemodynamic characteristics late postoperatively, we performed a multivariable analysis to determine which clinical and echocardiographic factors were associated with or protective against the progression of TR. In this analysis, the performance of a Maze procedure was identified as being protective against the progression of TR (odds ratio 0.09). Larger left atrial size, higher pulmonary artery pressure, worsened right ventricular function, and increased right ventricular dilatation were not significant predictive factors for the progression of TR in this study. Fig 2. Tricuspid regurgitation (TR) grade. Change in grade of tricuspid regurgitation for both the Maze (solid circles) and no-maze (solid boxes) groups during the preoperative (Preop), postoperative (Postop), and follow-up (FU) periods. Comment In this study, the restoration of sinus rhythm after addition of the Maze procedure during mitral valve surgery

4 Ann Thorac Surg STULAK ET AL 2008;86:40 5 MAZE PROCEDURE AF CORRECTION HALTS TR PROGRESS was associated with significantly decreased progression of late TR in comparison with patients who continued to have AF postoperatively. Further, in a multivariable model, the performance of the Maze procedure was identified as the only significant predictive factor in halting the progression of TR, thus implicating AF as the most significant factor contributing to the progression of TR in these patients. It is reported that the mechanism of TR regression after correction of mitral valve pathology is related to a decrease in right ventricular pressure and volume overload [5]. Several studies have shown, however, that despite relief of this pressure and volume back-up through the pulmonary vasculature by correction of mitral valve pathology, TR can worsen in late follow-up. Matsuyama and colleagues [6] identified several factors that were predictive of late TR after mitral valve surgery, including preoperative TR 2, persistent atrial fibrillation, and an enlarged left atrium. In the present study, we found that left atrial size continued to increase in patients who had persistent AF during follow-up when compared with patients who converted to sinus rhythm after the Maze procedure. Thus, it is not surprising that TR continued to progress in the no-maze group, as patients in this group harbored two significant risk factors (AF and enlarged left atrium) for the development of late TR after mitral valve surgery. It is known that the addition of the Maze procedure greatly increases the chance that patient will regain sinus rhythm after mitral valve surgery [15, 16], and as such, will offer patients the best chance to avoid progression of TR. There are several possible mechanisms by which persistent AF leads to progression of tricuspid valve insufficiency. In our patients, persistent AF was associated with higher pulmonary artery pressure late after operation in comparison with patients who had sinus rhythm who had undergone the Maze procedure. Kim and associates [17] first showed the positive effect of the Maze procedure in preventing progression of late TR after left-sided valve surgery, and these authors speculated that because left atrial dilation is frequently associated with an left atrial pressure elevation, associated pulmonary arteriolar constriction may lead to increased right ventricular afterload, right-sided chamber enlargement, and resultant TR. Patients in our study who had concomitant Maze procedure and remained in sinus rhythm had a decrease in left atrial size and a significant decrease in systolic pulmonary arterial pressures at late follow-up. Further, patients with persistent AF were more likely to have worsening right ventricular dilatation and right ventricular function in follow-up compared with patients in sinus rhythm. Atrial fibrillation is known to cause remodeling (enlargement) of the left atrium independent of left atrial pressure, and it is likely that the right atrium is affected in a similar fashion. In a study of elderly patients, Yamasaki and colleagues [18] reported that severe functional TR occurs with long-standing atrial fibrillation and causes right-sided heart failure. The authors speculated that TR is caused by loss of tricuspid valve systolic coaptation due to tricuspid annular dilation that, in turn, was caused by right atrial dilatation. There was a significant decrease in systolic pulmonary artery pressure in the Maze patients after sinus rhythm was restored, and it might be argued that this hemodynamic effect halted the progression of TR rather than restoration of sinus rhythm. But in the multivariable analysis increasing pulmonary artery pressure was not identified as a significant predictive factor in the progression of TR. This study did not address two other patient groups: patients who underwent mitral valve surgery without a Maze procedure and had spontaneous restoration of sinus rhythm; and patients who underwent combined mitral valve surgery and Maze procedure but remained in AF. We suspect that spontaneous restoration of sinus rhythm would likely yield similar results to the ones found in patients who underwent the Maze procedure in this study. The method used to attain sinus rhythm is most likely not the important factor, but we chose to study patients who had the Maze procedure to investigate whether beneficial effects of sinus rhythm would be seen in patients who had atrial surgery. There is debate as to the effectiveness of atrial contraction in patients after Maze operations, but our study shows that the procedure does prevent atrial remodeling and associated functional tricuspid regurgitation. Patients who remain in AF after Maze operations will likely be predisposed to further atrial dilatation, and TR as was seen in patients with persistent AF in this study. There was an important difference in the mitral valve operations performed in the two groups of patients in this study, and it is possible that mitral valve replacement of the underlying pathology in these patients contributed to the subsequent progression of TR. Seven patients with persistent AF (no Maze) who had mitral valve replacement had rheumatic heart disease compared with only 2 in the Maze group. However, all patients had functional TR, and organic disease was not seen in any patient on late follow-up echocardiography. Again, in the multivariable model, persistent AF was the only clinical or hemodynamic variable associated with late TR progression. The decision whether to address functional tricuspid regurgitation at the time of mitral valve surgery is important because of the undesired late clinical sequelae that can develop in these patients [4]. It is possible that an aggressive strategy to correct minor degrees of TR at the time of mitral valve repair or replacement would have prevented late TR in patients with AF. Indeed, Dreyfus and associates [19] advocate concomitant tricuspid valve annuloplasty if the tricuspid annular diameter is greater than twice the normal size ( 70 mm) regardless of the grade of regurgitation. Findings in this study suggest that this protocol may not be necessary for patients who maintain sinus rhythm postoperatively or will be converted to sinus rhythm by the Maze operation. There are several limitations of this retrospective study. Patients were selected based on postoperative rhythm, and despite case-matching, there were some baseline differences in the two groups. Secondly, fol- 43

5 44 STULAK ET AL Ann Thorac Surg MAZE PROCEDURE AF CORRECTION HALTS TR PROGRESS 2008;86:40 5 low-up of the patients in the Maze group was somewhat shorter than of patients who had AF, and it can be argued that the longer follow-up period accounted in part for the difference in TR progression. We believe this is unlikely because there were no time-related trends identified in TR. In addition, we chose to convert TR grade into a continuous variable. Because difference in TR grade is often not an equal amount, this analysis lends itself to errors inherent to this. To confirm the findings in changes in TR grade, we examined amount of TR progression and TR regression. In summary, tricuspid valve regurgitation is common among patients undergoing mitral valve surgery, and persistent or progressive TR has an important negative impact on late functional status. Atrial fibrillation, which is common in the setting of mitral valve pathology, is associated with atrial remodeling, which often leads to progressive tricuspid valve regurgitation. Ablation of AF at the time of mitral valve repair or replacement appears to prevent progression of TR and should improve late functional results. We would like to thank Donna Stucky for her assistance in editorial review. References 1. Czer LS, Maurer G, Bolger A, et al. Tricuspid valve repair. Operative and follow-up evaluation by Doppler color flow mapping. J Thorac Cardiovasc Surg 1989;98: Goldman ME, Guarino T, Fuster V, Mindich B. The necessity for tricuspid valve repair can be determined intraoperatively by two-dimensional echocardiography. J Thorac Cardiovasc Surg 1987;94: Cohen SR, Sell JE, McIntosh CL, Clark RE. Tricuspid regurgitation in patients with acquired chronic pure mitral regurgitation. II. Non-operative management, tricuspid valve annuloplasty and tricuspid valve replacement. J Thorac Cardiovasc Surg 1987;94: Porter A, Shapira Y, Wurzel M, et al. Tricuspid regurgitation late after mitral valve replacement: clinical and echocardiographic evaluation. J Heart Valve Dis 1999;8: Braunwald NS, Ross J Jr, Morrow AG. Conservative management of tricuspid regurgitation in patients undergoing mitral valve replacement. Circulation 1967;35(Suppl):I Matsuyama K, Matsumoto M, Sugita T, Nishizawa J, Tokuda Y, Matsuo T. Predictors of residual tricuspid regurgitation after mitral valve surgery. Ann Thorac Surg 2003;75: King MR, Schaff HV, Danielson GK, et al. Surgery for tricuspid regurgitation late after mitral valve replacement. Circulation 1984;70: Baughman KL, Kallman CH, Yurchak PM, Daggett WM, Buckley MJ. Predictors of survival after mitral valve surgery. Am J Cardiol 1984;54: Hornick P, Harris PA, Taylor KM. Tricuspid valve replacement subsequent to previous open heart surgery. J Heart Valve Dis 1996;5: 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: Jessurun ER, van Hemel NM, Kelder JC, et al. Mitral valve surgery and atrial fibrillation: is atrial fibrillation surgery also needed? Eur J Cardiothorac Surg 2000;17: Cox JL, Boineau JP, Schuessler RB, Kater KM, Lappas DG. Five-year experience with the Maze procedure for atrial fibrillation. Ann Thorac Surg 1993;56: Bando K, Kobayashi J, Kosakai Y, et al. Impact of Cox-Maze procedure on outcome in patients with atrial fibrillation and mitral valve disease. J Thorac Cardiovasc Surg 2002;124: McCarthy PM, Cosgrove DM III, Castle LW, White RD, Klein AL. Combined treatment of mitral regurgitation and atrial fibrillation with valvuloplasty and the Maze procedure. Am J Cardiol 1993;71: Handa N, Schaff HV, Morris JJ, Anderson BJ, Kopecky SL, Enriquez-Sarano M. Outcome of valve repair and the Cox Maze procedure for mitral regurgitation and associated atrial fibrillation. J Thorac Cardiovasc Surg 1999;118: Stulak JM, Sundt TM III, Dearani JA, Daly RC, Orszulak TA, Schaff HV. Ten-year experience with the Cox-Maze procedure for atrial fibrillation: how do we define success? Ann Thorac Surg 2007;83: Kim HK, Kim YJ, Kim KI, et al. Impact of the Maze operation combined with left-sided valve surgery on the change in tricuspid regurgitation over time. Circulation 2005;112(Suppl):I Yamasaki N, Kondo F, Kubo T, et al. Severe tricuspid regurgitation in the aged: atrial remodeling associated with long-standing atrial fibrillation. J Cardiol 2006;48: Dreyfus GD, Corbi PJ, Chan KM, Bahrami T. Secondary tricuspid regurgitation or dilatation: which should be the criteria for surgical repair? Ann Thorac Surg 2005;79: DISCUSSION DR WALTER MERRILL (Cincinnati, OH): This is a very interesting paper, and I congratulate the authors on their excellent work. There were a couple of points that I thought might deserve some further discussion. I noticed that in the patients who had the Maze procedure, the majority of them, 31 of 33, had a repair of the mitral valve, whereas in those who did not have a Maze procedure, repair was less frequent, 14 of 33. So obviously, there were more mitral valve replacements in the group that did not have a Maze. That may or may not have some influence on the late results, and perhaps you could comment on that. Also, the length of follow-up was a little bit different in that the patients who did not have a Maze had a longer period of follow-up as compared with those who had the Maze procedure. Possibly, with longer follow-up, there is more of a chance for the patients to go into atrial fibrillation later, and maybe you could tell us a bit more about that. Another point of interest is that some of the patients who had a successful mitral valve procedure and a Maze and were in normal sinus rhythm still had progression of tricuspid regurgitation, and I wonder if you might speculate as to the reasons why that was the case. And then finally, do you all still advocate doing nothing about the tricuspid regurgitation even if you successfully repair the mitral valve and establish sinus rhythm? Thank you. DR STULAK: Thank you, Dr Merrill. So your first question is why there were more replacements in the no-maze group. This was accounted for. The etiology of the mitral valve disease in that group was rheumatic in 8 patients, and of the patients in

6 Ann Thorac Surg STULAK ET AL 2008;86:40 5 MAZE PROCEDURE AF CORRECTION HALTS TR PROGRESS that group, there were more patients with mitral annular calcification. Now, you might argue that the presence of mitral valve etiology of rheumatic might have contributed to the worsening tricuspid valve. We are not sure that is the case because it didn t come out in a multivariable analysis, and also all patients with any echocardiographic evidence of organic tricuspid valve disease were excluded. So more rheumatics in that group and also more patients with mitral annular calcifications. There were 3 patients where a repair was attempted and the surgeon was not happy with the result, and then they just went right to replacement. So I think that accounts for the difference. There was a longer follow-up in the no-maze group, and this was not statistically significant, but you raise a good question: would it be clinically relevant, and if those Maze patients were allowed to go out 2 more years, would there be more atrial fibrillation? That is possible. We did look at time-related events of TR, and there was no time-related difference between the tricuspid valve regurgitation, but that is a good point. And then some of the Maze patients did progress despite being in sinus rhythm. There are many factors other than atrial fibrillation that causes progression of TR, and some that have been identified are preoperative TR grade greater than 2, which some of these patients did have, an enlarged left atrium preoperatively as well, which some of these patients also did have. So even though the Maze procedure does prevent atrial remodeling and halts the progression, we do acknowledge that some of these patients may go on to progress. And then secondly, what is our approach to secondary tricuspid regurgitation. If there is severe tricuspid regurgitation, we will fix it; if there is any evidence of right ventricular dysfunction or dilatation, we will fix it; and a lot of times if the etiology of the mitral valve disease is rheumatic and there is moderate tricuspid regurgitation, then the surgeon will choose to fix it. DR STULAK: We did not, and that is a great point. Tricuspid annular diameter has been implicated as actually a stronger predictor of progression of TR in patients rather than tricuspid regurgitation, because you can have tricuspid annular dilatation without tricuspid regurgitation. That is a great point, and we didn t address that here. DR HOOSHANG BOLOOKI (Miami, FL): Congratulations for a fine presentation of a very good study. The three components mitral valve, tricuspid valve, and atrial fibrillation go together, and it seems to me, based on our experience, with repairing mitral and doing Maze procedure for atrial fibrillation and repair of the tricuspid valve work well. If mitral repair does not hold and recurs in a few years, the chances of recurrence of atrial fibrillation and then tricuspid insufficiency are high. I have one question: were there mitral valves that were well repaired but on late follow-up, the tricuspid valve independently developed insufficiency or only atrial fibrillation suddenly recurred? Do you see the same, as I am indicating, does the mitral valve repair become the mainstay of the three components of this one operation that you are talking about? DR STULAK: So if I understand your question correctly, all patients here had similar reduction in mitral regurgitation, and these repairs were durable. So at follow-up, there was no difference in mitral regurgitation among these patients. DR BOLOOKI: None of them got worse, in other words? DR STULAK: Correct. The grades of mitral regurgitation at last follow-up were identical between groups. So the only variable between the groups is atrial fibrillation. 45 DR RICHARD LEE (Chicago, IL): I enjoyed your presentation, and I had one question. One additional fact that has been known to predict progression of tricuspid regurgitation is tricuspid annulus size even without tricuspid regurgitation at the time. Did you have a moment to look at the tricuspid annulus size in each of the groups and see if there is a difference or not? DR BOLOOKI: So atrial fibrillation independently came on? DR STULAK: Correct. DR BOLOOKI: Thank you very much.