Abstract
Objective
To determine the incidence and effects of tricuspid regurgitation (TR) after surgery for mitral valve (MV) repair for mitral regurgitation (MR) due to degenerative disease.
Patients and Methods
We examined 1171 patients who had MV repair and were followed prospectively with periodical clinical and echocardiographic assessments during a mean of 9.1 ± 5.3 years. Patients' mean age was 58.2 ± 12.7 years, and 70.5% were men. Preoperatively, 44.6% were in functional classes III and IV, 20.1% had atrial fibrillation, and 34.2% had ejection fraction <60%. In addition to MV repair, 13.8% had coronary artery bypass, 11.4% had the maze procedure, and 4.7% had tricuspid annuloplasty.
Results
Moderate and severe TR was present in 138 patients before surgery and associated with older age, preoperative atrial fibrillation, preoperative congestive heart failure, congenital heart septal defects, lower preoperative left ventricular ejection fraction, and female sex by multivariable analysis. TR resolved postoperatively but recurrent or new isolated TR occurred in 45 patients postoperatively (13.6% at 15 years in all patients). Factors associated with isolated postoperative TR by multivariable analysis included older age at operation, unrepaired preoperative moderate/severe TR, and the development of postoperative MR. Patients with preoperative TR had reduced long-term survival and tricuspid annuloplasty did not restore lifespan.
Conclusions
Preoperative TR in patients with MR due to degenerative diseases was associated with longstanding MV disease and adversely affected long-term survival after MV repair. New postoperative TR was uncommon. The findings of this study are compelling reasons to repair the MV before the development of TR.
Key Words
Abbreviations and Acronyms:
AATS (American Association for Thoracic Surgery), AF (atrial fibrillation), CI (confidence interval), HR (heart ratio), MR (mitral regurgitation), MV (mitral valve), OR (odds ratio), TA (tricuspid annuloplasty), TR (tricuspid regurgitation), TV (tricuspid valve)
Patients' survival after mitral valve repair.
Central Message
Functional tricuspid regurgitation after mitral valve repair for degenerative diseases of the mitral valve was uncommon in this long-term study.
Perspective
Functional tricuspid regurgitation in patients with mitral regurgitation due to degenerative diseases was associated with older age, chronic atrial fibrillation, advanced functional class, congenital heart defects, impaired left ventricular function, and female sex. These factors adversely affected long-term survival after valve repair and concomitant tricuspid annuloplasty did not restore lifespan.
See Editorial page 108.
This study was prompted by a presentation by Joanna Chikwe at the 2015 Annual Meeting of The American Association for Thoracic Surgery (AATS) in Seattle, where she reported on a series of 645 patients who had mitral valve (MV) repair for degenerative disease and 419 (65%) of them had concomitant tricuspid annuloplasty (TA) because of moderate tricuspid regurgitation (TR), tricuspid annular diameter ≥40 mm by echocardiography, or direct inspection of the tricuspid valve (TV) in equivocal cases.
1
Because 83% of the patients who TA had mild, trace, or no TR, the mean diameter of the tricuspid annulus was 38.8 mm, and more than one half of all patients had advanced myxomatous degeneration of the MV (which often involves the TV too), we argued that the rate of TA was excessively high and unnecessary. Robert Dion clearly disagreed with our views and gave the indications when TA should be combined with MV surgery in a recent editorial.2
A brief review of our data confirmed our views expressed at the AATS meeting.3
In another editorial in the Journal, Richard Shemin put things in perspective and correctly stated that we need more evidence and less personal opinions on issues as perplexing as functional TR.4
This study examined the incidence of TR and its consequences in a large cohort of patients who had MV repair for mitral regurgitation (MR) due to degenerative diseases during the last 3 decades. When we first began a program of MV repair for MR in mid-1980s, we overlooked the TV unless the right side of the heart was dilated grossly at surgery or the patient had had evidence of right-side heart failure preoperatively because we believed that the TR would resolve after correction of the left-side lesion. Moreover, in reviewing the echocardiographic reports from studies done in the 1980s and 1990s, we found that the TV also was overlooked frequently by the echocardiographers because their written reports contained no information on the functional status of this valve or simply stated that the right-sided valves were normal without quantification of TR. We observed, however, that a few patients developed transient right-side failure during the first weeks after MV repair and that these patients often had echocardiographic evidence of moderate or severe TR before surgery; as a result, we increased the rate of TAs we performed over time.
Patients and Methods
Isolated MV repair for MR due to degenerative diseases of the MV was performed in 1171 patients by one surgeon from 1985 to 2010. The clinical outcomes of MV repair on these patients were published recently in 2 separate reports, but the issue of concomitant TR was not addressed.
5
, 6
We used the database from those 2 reports to design this study. Our patients were followed by the referring cardiologists, and both patients and cardiologists were contacted by our research personnel periodically (yearly during the first 2-3 years and approximately every 2-3 years thereafter), and an echocardiogram was requested to assess MV and ventricular function.The echocardiograms were read by various cardiologists, who sent us a written report. Valve regurgitation was recorded as none, trivial, mild, moderate, and severe (if a report read “mild to moderate,” it was counted as moderate and if read “moderate to severe,” it was counted as “severe”) and entered in a dedicated database. Most echocardiographic reports from 1980s and early 1990s had detailed information on the aortic and MVs but scant information on TV, and sometimes read “normal right-side valves.” In 119 patients who had no preoperative echocardiographic data on the TV, we used the TR grading obtained in the operating room (transepicardial echocardiography in the first couple of years of the study and transesophageal since 1988). Postoperative studies that reported the TV as “normal” but contained no information of the severity TR were excluded from analysis. Since the publication of the guidelines by American Society of Echocardiography in 2003, all reports contained detailed information on all 4 heart valves.
7
The diameter of the tricuspid annulus was only sporadically measured and not entered in the datasets.Adverse events were reported according to the guidelines set by cardiac surgical societies.
8
The cause of death was determined by hospital chart review, death certificates, or information from the physician who was caring for the patient at that time. Clinical follow-up was complete in 98.1% (defined as patient having a terminal event or being contacted within 2 years from our previous reports on MV repair5
, 6
). Echocardiography was performed in all patients in the operating room and before discharge from hospital. Echocardiographic data were available in 95.7% of patients alive at the last follow-up. The overall mean follow-up was 9.1 ± 5.3 years, but it was shorter at 6.9 ± 4.9 years for patients who had TA combined with MV repair for the aforementioned reasons. Twelve patients with TR had leaflet prolapse and were included in the group as functional TR. This study was approved by the Review Ethics Board of the University Health Network and patients' consent waived.Statistical Analysis
Data are presented as means with standard deviation, median with interquartile range (25th and 75th percentile), and frequencies as appropriate. Comparisons of patients' clinical profile and perioperative outcomes (Tables 1 and 2) were performed by either the Fisher exact test, Wilcoxon 2-sample test, or Student t test assuming unequal variance between groups, as appropriate. Visual inspection of the distribution for continuous variables was used to assess normality. Factors associated with preoperative moderate/severe TR and undergoing concomitant TA were modeled with a logistic regression models (see below for model building strategy). Recognizing that patients included in this study were at risk of multiple postoperative outcomes at the same time, a nonparametric competing risk model was developed. In this model, patients could transition to 1 of 5 possible endpoints, whichever occurred first. These endpoints are listed to follow, with the remaining of the patients being alive and free from reoperation or moderate/severe MR or TR:
- •reoperation for isolated MR;
- •reoperation for isolated TR;
- •reoperation for concomitant MR/TR;
- •reoperation for reasons other than MR or TR; or
- •death before reoperation.
Table 1Patients' clinical profiles
| Variable | All patients | MV repair | MV + TV repair | P value |
|---|---|---|---|---|
| Number of patients | 1171 | 1116 | 55 | |
| Age at surgery, y, average ± SD | 58.2 ± 12.7 | 57.9 ± 12.6 | 65.5 ± 13.4 | <.001 |
| Sex (male) | 826 (70.5) | 798 (71.5) | 28 (50.9) | .002 |
| Previous surgical cardiac intervention | 22 (1.9) | 20 (1.8) | 2 (3.6) | .28 |
| Urgent operation | 84 (7.3) | 78 (7.2) | 6 (10.9) | .66 |
| NYHA class | <.001 | |||
| Class I | 224 (19.1) | 220 (19.7) | 4 (7.3) | .02 |
| Class II | 437 (37.3) | 426 (38.2) | 11 (20.0) | .007 |
| Class III | 412 (35.2) | 384 (34.4) | 28 (50.9) | .01 |
| Class IV | 98 (8.4) | 86 (7.7) | 12 (21.8) | .001 |
| Left ventricular ejection fraction | <.001 | |||
| ≥60% | 771 (65.8) | 751 (67.2) | 20 (37) | <.001 |
| 40%-59% | 349 (29.8) | 322 (28.9) | 27 (50) | .002 |
| 20%-39% | 50 (4.3) | 43 (3.9) | 7 (13) | <.001 |
| <20% | 1 (0.1) | 1 (0.1) | 0 | >.99 |
| Atrial fibrillation or flutter | 235 (20.1) | 206 (18.5) | 29 (52.7) | <.001 |
| Complete heart block | 12 (1.0) | 7 (0.6) | 5 (9.1) | <.001 |
| History of congestive heart failure | 539 (46.0) | 495 (44.4) | 44 (80.0) | <.001 |
| Associated diseases | ||||
| Diabetes | 50 (4.3) | 48 (4.3) | 2 (3.6) | >.99 |
| Hypertension | 344 (29.4) | 330 (29.6) | 14 (25.5) | .65 |
| Hyperlipidemia | 282 (24.1) | 270 (24.2) | 12 (21.8) | .75 |
| Family history of coronary artery disease | 445 (38.9) | 434 (39) | 20 (36.4) | .78 |
| Smoking | 508 (43.5) | 478 (43) | 30 (54.6) | .10 |
| Chronic obstructive lung disease (FEV1 < 1) | 34 (2.9) | 33 (3) | 1 (1.8) | >.99 |
| Previous stroke/transient ischemic attack | 63 (5.4) | 58 (5.2) | 5 (9.1) | .22 |
| Peripheral vascular disease | 8 (0.7) | 8 (0.7) | 0 | >.99 |
| Marfan syndrome | 16 (2) | 16 (2.1) | 0 | >.99 |
| Renal failure | 3 (0.3) | 3 (0.3) | 0 | >.99 |
| Creatinine >150 μmol/L | 6 (0.7) | 6 (0.7) | 0 | >.99 |
| Infective endocarditis | ||||
| Active | 5 (0.4) | 5 (0.5) | 0 | >.99 |
| Remote | 95 (8.1) | 91 (8.2) | 4 (7.3) | .80 |
| Congenital heart defects (ASD, VSD) | 36 (3.1) | 30 (2.7) | 6 (10.9) | .005 |
| Ascending aorta dilatation | 11 (0.9) | 9 (0.8) | 2 (3.6) | .09 |
| Coronary artery disease | 162 (13.8) | 155 (13.9) | 7 (12.7) | .64 |
| Echocardiographic data | ||||
| TV regurgitation | <.001 | |||
| None/trivial | 513 (43.8) | 513 (45.9) | 0 | <.001 |
| Mild | 520 (44.4) | 520 (46.6) | 0 | <.001 |
| Moderate | 99 (8.4) | 73 (6.5) | 26 (47.3) | <.001 |
| Severe | 39 (3.3) | 10 (0.9) | 29 (52.7) | <.001 |
| Tricuspid valve pathology | <.001 | |||
| Functional tricuspid regurgitation | 102 (8.7) | 59 (5.2) | 43 (78.1) | <.001 |
| Tricuspid regurgitation with leaflet prolapse | 12 (1) | 0 | 12 (20.8) | <.001 |
| Mitral regurgitation | >.99 | |||
| Moderate | 62 (5.3) | 59 (5.2) | 3 (4.9) | >.99 |
| Severe | 1009 (94.7) | 1057 (94.7) | 52 (95) | >.99 |
| MV prolapse | ||||
| None (annular dilatation) | 18 (1.5) | 15 (1.3) | 3 (4.9) | .06 |
| Posterior leaflet | 460 (39.2) | 438 (39.2) | 22 (40) | .05 |
| Anterior leaflet | 214 (18.2) | 208 (18.4) | 6 (10.9) | .21 |
| Bileaflet | 479 (40.9) | 455 (40.7) | 24 (43.6) | >.99 |
| MV pathology | ||||
| Advanced myxomatous degeneration | 254 (21.7) | 245 (21.9) | 9 (16.3) | .93 |
| Dystrophic calcification of the mitral annulus | 32 (2.7) | 30 (2.7) | 2 (3.6) | >.99 |
| Operative data | ||||
| MV repair | 1171 (100) | 1116 (100) | 55 (100) | >.99 |
| Mitral annulus reconstruction | 32 (2.7) | 30 (2.7) | 2 (3.6) | >.99 |
| Mitral annuloplasty | ||||
| Carpentier ring | 101 (8.6) | 98 (8.7) | 3 (5.4) | .62 |
| Duran ring | 152 (12.9) | 145 (13) | 7 (12.7) | >.99 |
| Cosgrove band | 509 (43.4) | 490 (43.9) | 19 (34.6) | .17 |
| Simplici-T band | 337 (28.8) | 312 (28) | 25 (45.5) | .009 |
| No ring | 72 (6.2%) | 71 (6.4) | 1 (1.8) | .25 |
| Tricuspid annuloplasty | ||||
| DeVega | – | – | 28 (50.9) | |
| Simplici-T band | – | – | 22 (40) | |
| Carpentier ring | – | – | 5 (9) | |
| Closure of atrial or ventricular septal defects | 36 (3.1) | 30 (2) | 6 (10.9) | .005 |
| Ascending aorta replacement | 11 (0.9) | 9 (0.8) | 2 (3.6) | .09 |
| Maze procedure | 133 (11.4) | 111 (10) | 22 (40) | <.001 |
| Coronary artery bypass graft | 162 (13.8) | 155 (13.9) | 7 (12.7) | .64 |
| Cardiopulmonary bypass time, min | 82 ± 26 | 81 ± 25 | 107 ± 29 | <.001 |
| Aortic clamping time, min | 65 ± 23 | 64 ± 22 | 84 ± 23 | <.001 |
Percentages are shown in parentheses. MV, Mitral valve; TV, tricuspid valve; SD, standard deviation; NYHA, New York Heart Association; FEV1, forced expiratory volume per second; ASD, atrial septal defect; VSD, ventricular septal defect.
Table 2Perioperative outcomes
| All patients | MV repair | MV + TV repair | P value | |
|---|---|---|---|---|
| Ventilation time, h, median (25th-75th) | 5 (4-7) | 5 (4-7) | 7 (5-11) | .04 |
| ICU stay, h, median (25th-75th) | 24 (21-42) | 24 (21-28) | 46 (24-74) | .002 |
| Hospital stay, d, median (25th-75th) | 7 (6-9) | 7 (6-9) | 8 (7-11) | .03 |
| Intra- or postoperative IABP | 14 (1.2) | 10 (0.9) | 4 (7.3) | .003 |
| Reoperations | ||||
| Any reoperation | 56 (4.8) | 54 (4.8) | 2 (3.6) | >.99 |
| Bleeding/tamponade | 49 (4.2) | 48 (5.4) | 1 (1.8) | .72 |
| Shock/arrest + bleeding | 1 | 1 | 0 | >.99 |
| Sternal infection | 1 | 1 | 0 | >.99 |
| Sternal dehiscence | 1 | 1 | 0 | >.99 |
| Redo mitral valve surgery | 2 | 1 | 1 | .09 |
| Other | 2 (0.2) | 2 (0.2) | 0 | >.99 |
| Perioperative myocardial infarction | 7 (0.6) | 7 (0.6) | 0 | >.99 |
| Low output syndrome | 39 (3.3) | 33 (3.0) | 6 (10.9) | .008 |
| Dialysis required | 4 (0.3) | 4 (0.4) | 0 | >.99 |
| Insertion of permanent pacemaker | 39 (3.3) | 32 (2.9) | 7 (12.7) | .002 |
| New postoperative atrial fibrillation | 239 (20.4) | 232 (20.8) | 7 (12.7) | .17 |
| Pulmonary complications | 113 (9.7) | 108 (9.7) | 5 (9.1) | >.99 |
| Seizures | 9 (0.8) | 9 (0.8) | 0 | >.99 |
| Transient ischemic attack | 6 (0.5) | 6 (0.5) | 0 | >.99 |
| Stroke | 7 (0.6) | 7 (0.6) | 0 | >.99 |
| Sepsis | 4 (0.3) | 4 (0.4) | 0 | >.99 |
| Operative mortality | 6 (0.5) | 4 (0.4) | 2 (3.6) | .03 |
MV, Mitral valve; TV, tricuspid valve; ICU, intensive care unit; IABP, intra-aortic balloon pump.
Cox proportional hazard was used to identify factors associated with postoperative mortality. For both the binary and time-dependent outcomes, the strategy for model building was similar. For the first time, outcome-specific univariable regression models were used to screen potential risk factors (Table E1), and risk factors with univariable P values <.20 were then included in a bootstrap resampling algorithm (1000 random resamples). Variables with a high reliability (ie, selection in >50% of the samples) were then included in a multivariable regression models with backward selection of variables to obtain a final model. TR over time was modeled with the use of longitudinal mixed models, with backward selection of variables that were statistically significant at the univariable level to obtain a final model. Mean imputation was used for the rare missing data points in potential predictor variables. All statistical analyses were performed with SAS v9.3 (SAS Institute, Cary, NC).
Results
Preoperative and Intraoperative Characteristics and Perioperative Outcomes
A total of 1171 patients were included in this study, of whom 1116 (95.3%) underwent isolated MV repair and 55 (4.7%) underwent combined MV repair and TA. Table 1 shows the clinical profile and operations performed in all patients and stratified by whether the patient underwent concomitant TA. Table 2 shows the perioperative outcomes. Before surgery, 99 (8.5%) patients had moderate TR, and 39 (3.3%) had severe TR. All patients who underwent TA had moderate or severe TR preoperatively. Factors associated with preoperative moderate or severe TR are listed in Table 3 and included preoperative atrial fibrillation (AF), older age at operation, advanced functional class/congestive heart failure, congenital heart disease, lower preoperative ejection fraction, and female sex. Interestingly, odds of preoperative TR were not associated with year of surgery either in univariable analyses (odds ratio [OR], 0.98/year, 95% confidence interval [95% CI], 0.96-1.01, P =.24) or when added to the multivariable model described previously (OR, 1.02/year, 95% CI, 0.99-1.05, P =.26).
Table 3Multivariable factors associated with preoperative moderate/severe tricuspid valve regurgitation
| Factor | Reliability | M.OR (LCL-UCL) | P value |
|---|---|---|---|
| Preoperative atrial fibrillation | 99.0% | 2.444 (1.589-3.761) | <.001 |
| Age at operation (per 5 y) | 98.6% | 1.200 (1.093-1.319) | <.001 |
| Advanced functional class/CHF | 94.0% | 2.051 (1.280-3.287) | .003 |
| Congenital heart disease | 93.2% | 5.114 (2.139-12.23) | <.001 |
| Preoperative LVEF (per level) | 82.2% | 1.585 (1.142-2.201) | .006 |
| Sex = female | 70.4% | 1.599 (1.060-2.412) | .03 |
M.OR, Multivariable odds ratio; LCL, lower confidence limit (95%); UCL, upper confidence limit (95%); CHF, congestive heart failure; LVEF, left ventricular ejection fraction.
Not all patients who had moderate or severe preoperative TR had concomitant TA, and as shown in Figure 1, and the probability of undergoing concomitant TA in those patients increased with later year of surgery (hazard ratio [HR], 1.10/year, 95% CI, 1.04-1.16, P <.001), indicating increasing awareness of a potential TR problem over time. In addition to year of operation, other factors associated with undergoing concomitant TA are listed in Table 4. Those factors included greater preoperative TR grade, lower preoperative ejection fraction, congenital heart septal disease, greater preoperative functional class, and older age at operation. Early postoperative outcome stratified by isolated MV repair versus MV repair with concomitant TA are listed in Table 2 and showed an expectedly worsened postoperative profile in patients with concomitant MV repair and TA.
Figure 1Increased probability of tricuspid annuloplasty over the years in patients with moderate and severe tricuspid regurgitation. OR, Odds ratio.
Table 4Multivariable factors associated with concomitant tricuspid annuloplasty
| Factor | Reliability | M.OR (LCL-UCL) | P value |
|---|---|---|---|
| Preoperative TR grade – per grade | 100.0% | 6.123 (4.062-9.229) | <.001 |
| Year of operation | 100.0% | 1.149 (1.088-1.214) | <.001 |
| Preoperative LV grade – per class | 98.6% | 2.610 (1.559-4.369) | <.001 |
| Congenital heart disease | 73.6% | 5.135 (1.429-18.45) | .01 |
| Preoperative NYHA – per class | 72.4% | 1.689 (1.056-2.700) | .03 |
| Age at operation (per 5 y) | 65.2% | 1.181 (1.009-1.382) | .04 |
M.OR, Multivariable odds ratio; LCL, lower confidence limit (95%); UCL, upper confidence limit (95%); TR, tricuspid regurgitation; LV, left ventricle; NYHA, New York Heart Association.
Long-Term Outcomes
Figure 2 shows long-term competing outcomes separately for all patients (upper panel), for patients without or with moderate/severe preoperative TR or concomitant TA (middle and lower panels respectively). Prevalence at various time points of all competing outcomes over time since initial surgery is listed in Table 5. There were 6 early and 178 late deaths: 82 valve or cardiac-related, 93 noncardiac or valvular, and 3 unknown causes. The operative mortality was greater in patients who had TA (0.4% vs 3.6%, P =.03). There were 43 (3.7%) reoperations for recurrent MR (38 patients) or MR with TR (5 patients) but no reoperation for isolated TR, although severe TR was the principal cause of death in 2 elderly patients deemed “inoperable.” A total of 10 patients had reoperation for reasons other than MR or TR.
Figure 2Nonparametric prevalence of competing outcomes over time after mitral valve repair. Upper panel represents outcomes for the entire cohort, middle panel represents outcomes for patients without tricuspid regurgitation, and lower panel represents patients with preoperative moderate or severe tricuspid regurgitation, respectively. Dashed lines represent the standard error around the cumulative estimate incidence. N left, Number of patients at risk; % NO, percent alive and free from any adverse event; TR, tricuspid regurgitation; MR, mitral regurgitation.
Table 5Prevalence (with standard error) of outcomes in competing risk models stratified by preoperative tricuspid valve status and concomitant TA
| All patients | Recurrent Isolated MR | Recurrent Isolated TR | Recurrent MR/TR | Reoperation no MR/TR | Death without reoperation or MR/TR | Alive, no reoperation, no MR/TR |
|---|---|---|---|---|---|---|
| All patients | ||||||
| n | 38 | 45 | 11 | 10 | 160 | 907 |
| 1 y | 1.0 ± 0.3 | 0.1 ± 0.1 | 0.1 ± 0.1 | 0.1 ± 0.1 | 1.2 ± 0.3 | 97.5 ± 0.5 |
| 5 y | 2.0 ± 0.4 | 0.5 ± 0.2 | 0.1 ± 0.1 | 0.4 ± 0.2 | 4.3 ± 0.6 | 92.8 ± 0.8 |
| 10 y | 3.4 ± 0.6 | 1.8 ± 0.5 | 0.6 ± 0.3 | 0.6 ± 0.2 | 11.3 ± 1.1 | 82.2 ± 1.3 |
| 15 y | 4.2 ± 0.8 | 6.9 ± 1.2 | 2.0 ± 0.7 | 0.8 ± 0.3 | 20.3 ± 1.8 | 65.8 ± 2.2 |
| Patients without preoperative TR or concomitant TA | ||||||
| n | 33 | 31 | 5 | 9 | 136 | 825 |
| 1 y | 1.0 ± 0.3 | 0.1 ± 0.1 | 0.0 ± 0.0 | 0.1 ± 0.1 | 0.9 ± 0.3 | 98.0 ± 0.4 |
| 5 y | 1.9 ± 0.4 | 0.2 ± 0.1 | 0.0 ± 0.0 | 0.4 ± 0.2 | 3.9 ± 0.6 | 93.6 ± 0.8 |
| 10 y | 3.4 ± 0.6 | 1.3 ± 0.5 | 0.2 ± 0.2 | 0.7 ± 0.3 | 10.8 ± 1.2 | 83.5 ± 1.4 |
| 15 y | 3.9 ± 0.7 | 5.8 ± 1.3 | 0.9 ± 0.4 | 0.9 ± 0.4 | 19.5 ± 1.9 | 68.9 ± 2.3 |
| Patients with preoperative TR or concomitant TA | ||||||
| n | 5 | 14 | 6 | 1 | 24 | 82 |
| 1 y | 1.5 ± 1.1 | 0.0 ± 0.0 | 0.8 ± 0.8 | – | 3.8 ± 1.7 | 93.9 ± 2.1 |
| 5 y | 2.3 ± 1.3 | 2.5 ± 1.4 | 0.8 ± 0.8 | – | 7.7 ± 2.3 | 86.8 ± 3.0 |
| 10 y | 3.4 ± 1.7 | 7.1 ± 2.6 | 4.1 ± 2.0 | – | 14.9 ± 3.6 | 70.5 ± 4.7 |
| 15 y | 5.8 ± 2.9 | 14.9 ± 4.5 | 8.8 ± 3.8 | – | 25.3 ± 5.4 | 45.1 ± 6.7 |
MR, Mitral regurgitation; TR, tricuspid regurgitation.
Multivariable factors associated with postoperative mortality are listed in Table 6 and included older age at surgery, early year of operation, lower LV grade, preoperative hypertension, preoperative AF, postoperative (<1 month) AF, and postoperative (<1 month) pacemaker implantation. Removing the 2 postoperative risk factors from the model did not add or remove any preoperative risk factors from the final model. Interestingly, the presence of TR preoperatively was associated with increased odds of mortality in univariable analyses (HR, 1.99, 95% CI, 1.37-2.89, P <.001) but failed to meet the reliability threshold of 50% (reliability: 24.6%) and did not reach statistical significance when added to the multivariable model described previously (HR, 1.19, 95% CI, 0.81-1.76, P =.38).
Table 6Multivariable factors associated with postoperative isolated moderate/severe tricuspid valve regurgitation
| Factor | Reliability | M.HR (LCL-UCL) | P value |
|---|---|---|---|
| Age at operation (per 5 y) | 99.8% | 1.487 (1.239-1.786) | <.001 |
| Postoperative (<1 mo) TR grade | 98.2% | 2.842 (1.815-4.448) | <.001 |
| Insertion of permanent pacemaker | 95.4% | 7.428 (3.066-18.00) | <.001 |
| Preoperative atrial fibrillation | 83.2% | 2.200 (1.209-4.002) | .01 |
M.HR, Multivariable hazard ratio; LCL, lower confidence limit (95%); UCL, upper confidence limit (95%); TR, tricuspid regurgitation.
Incidence of Factors and Outcomes Associated with Postoperative TR
The development of TR after MV repair in patients without preoperative TR was rare. Moderate preoperative TR resolved in all but 3 patients and in 8 of 10 with severe preoperative TR, but 15 developed TR again during the follow-up. The incidence of moderate/severe TR over time is reported in Figure 3 with point estimates at 1, 5, 10, and 15 years and is provided in Table 7 along with stratification by preoperative TR and TA. Of note, <5% of postoperative TR was in the severe range. Factors associated with isolated postoperative TR by multivariable analysis are shown in Table 8 and included older age at operation, unrepaired preoperative moderate/severe TR, and the development of postoperative MR.
Figure 3Incidence of moderate/severe tricuspid regurgitation over time for all patients (top) and stratified by preoperative tricuspid valve regurgitation and concomitant tricuspid annuloplasty (bottom). Dotted lines represent the 95% confidence interval. TR, Tricuspid regurgitation; Preop, preoperative; TV, tricuspid valve.
Table 7Incidence of TR (moderate) over time (note: <5% of all TR was greater than moderate)
| Years | Overall | No preoperative TR | Preoperative TR without tricuspid annuloplasty | Preoperative TR with tricuspid annuloplasty |
|---|---|---|---|---|
| 1 | 4.3% (3.3%, 5.3%) | 1.7% (1.2%, 2.3%) | 45.5% (35.5%, 55.4%) | 7.8% (1.9%, 13.5%) |
| 5 | 6.0% (4.9%, 7.1%) | 3.0% (2.2%, 3.8%) | 51.9% (41.4%, 63.0%) | 6.1% (0.1%, 13.2%) |
| 10 | 9.1% (7.2%, 11.1%) | 5.9% (4.4%, 7.4%) | 59.8% (45.5%, 74.1%) | 4.4% (0.0%, 18.3%) |
| 15 | 13.6% (9.6%, 17.5%) | 11.5% (7.9%, 15.4%) | 67.1% (46.4%, 83.8%) | 3.1% (0.0%, 27.7%) |
Numbers in parentheses show 95% confidence interval. TR, Tricuspid regurgitation.
Table 8Multivariable factors associated with recurrent TR
| Factor | M.HR (LCL-UCL) | P value |
|---|---|---|
| Age at operation (per 10 y) | 2.13 (1.64-2.92) | <.001 |
| Preoperative TR/TA | ||
| No TR | Reference | |
| TR with TA | 2.96 (0.92-9.19) | .06 |
| TR without TA | 98.3 (39.7-344.4) | <.001 |
| Postoperative MR grade | ||
| Less than moderate | 0.20 (0.10-0.35) | <.001 |
TR, Tricuspid regurgitation; M.HR, multivariable hazard ratio; LCL, lower confidence limit (95%); UCL, upper confidence limit (95%); TA, tricuspid annuloplasty; MR, mitral regurgitation.
∗ Collected longitudinally.
Impact of Performing TA in Patients With Preoperative Moderate-Severe TR on Long-Term Outcomes
Moderate or severe preoperative TR was associated with an increased risk of postoperative mortality, but performance of TA did not reduce the risk (Figure 4). Eleven patients who had TA developed recurrent moderate TR during the follow-up (7 had had DeVega annuloplasty, 3 Simplici-T band, and 1 Carpentier ring).
Figure 4Survival over time stratified by preoperative tricuspid valve regurgitation and concomitant TA. TR, Tricuspid regurgitation; TA, tricuspid annuloplasty; HR, hazard ratio; CI, confidence interval.
Discussion
This study examined the incidence and effect of TR in a large cohort of patients who had MV repair for degenerative diseases. TR was addressed infrequently during the first half of this experience and only patients with evidence of right-side failure had TA. As we became more aware of TR, we began to correct TR at the time of MV repair with a consequent increase in the number of TA over time, as documented in Figure 1 and Table 4. We found that moderate or severe TR at the time of MV repair for degenerative diseases of the MV was associated with older age, impaired left ventricular function, AF, more advanced symptoms of congestive heart failure, congenital heart septal defects, and female sex by multivariable analysis (Table 3). This study showed that these patients probably should have TA at the time of MV repair but long-term survival is reduced because of the associated risk factors rather than the presence of TR. This is another compelling reason to correct MR before patients develop advanced symptoms of heart failure, left ventricular dysfunction, or AF.
As we stated during the discussion of the paper by Chikwe and colleagues
1
at the AATS meeting in 2015, the development of isolated TR after MV repair for degenerative disease was uncommon in our patients, and it was associated with advanced age at operation, unrepaired preoperative moderate/severe TR, and the development of postoperative MR as shown in Table 6. Forty-five patients developed isolated moderate or severe TR after MV repair, but 11 of them had had TA. In addition, 11 patients developed recurrent severe MR and moderate or severe TR. The estimated risk of developing isolated moderate or severe TR was 13.6% at 15 years after MV repair in all patients. In the study by Chikwe and colleagues,1
TR occurred in 9% of patients who did not have TA and in 3% of those who did only 5 years after surgery.Murashita and colleagues
9
reported an incidence of TR similar to ours after MV repair for degenerative diseases in a cohort of 479 patients who were followed for a mean of 7.5 years. They found that AF and the grade of TR before surgery were associated with the development of severe TR late postoperatively. This finding endorses the performance of TA in patients with moderate or severe TR. Yilmaz and colleagues10
from the Mayo Clinic examined the outcomes of 693 patients who had isolated MV repair for MR due to MV prolapse and 16% had moderate or severe TR before surgery. They excluded patients with coronary artery disease and congenital heart defects. They observed a significant reduction in the overall grade of TR within the first 3 years but a slight increase after 5 years. Female sex, preoperative AF, and diabetes mellitus were independent risk factors for increased TR with time.10
A more recent evaluation of this issue at that institution confirmed that MV repair as well as MV replacement reduced TR grade postoperatively and the development of new TR was uncommon.11
Although the number of patients who developed severe TR after MV repair was low in our study (less than 5% of recurrent TR was greater than moderate), the effect of severe TR was devastating, with a high rate of mortality at 1 year after diagnosis, which usually was made during an episode of heart failure. Topilsky and colleagues
12
from the Mayo Clinic examined the outcomes of 353 patients (mean age 70 years) with isolated TR and found that an effective regurgitant orifice ≥40 mm2 was associated with marked reduction in survival independent of all characteristics, right ventricular size or function, comorbidity, or pulmonary pressure and lower than expected in the general population. Other investigators have found that severe TR after implantation of transvenous pacemaker was associated with poor long-term survival.13
It is worth mentioning that insertion of permanent transvenous pacemaker after TA does not seem to increase the risk of TR.14
We have been performing the maze procedure for AF only since 1996 and that is the reason why only 133 of 235 with AF at the time of surgery had combined MV repair with maze. We could not establish a relationship between maze and postoperative TR, but it has been suggested that the maze procedure reduces the risk of postoperative TR if successful in eliminating AF.
15
Dreyfus and colleagues
16
introduced the concept that annular dilatation precedes the development of TR and a tricuspid annulus ≥70 mm in the arrested heart (the distance from the commissure between septal and anterior leaflets to the base of the posterior leaflet) should be an indication for TA regardless of TR grade. Nemoto and colleagues17
examined a large number of patients with no or trace TR, mild TR, and moderate or severe TR by echocardiography and computed tomography and found TV annular dilatation and atrial enlargement comprise early events in mild functional TR and atrial enlargement occurs before right ventricular dilation, which occurs late, when TR is severe. They concluded that atrial volume and TV annular dilatation are early and sensitive indicators of TR.17
The shape of the tricuspid annulus changed from elliptical to circular with severe TR.17
This study supports the concept that tricuspid annulus dilatation may precede functional TR. The European guidelines on the management of functional TR suggests that a preoperative echocardiographic TV annulus ≥40 mm is an indication for TA (class IIa, level C).
18
Several surgeons have incorporated this recommendation into their practice, and the rate of TA at the time of MV repair has soared to as high 65%.1
, 2
, 19
, 20
These investigators argued that the right ventricle remodels after TA in patients with dilated tricuspid annulus.1
, 19
The assessment of right ventricular size and function, however, is difficult and often inaccurate.21
Moreover, a recently published echocardiographic study Sordelli and colleagues22
from Milan, Italy, which included 3-dimensional imaging, failed to show a correlation between the diameters of the TV annulus (anteroposterior and septolateral) and the development of TR after MV repair for degenerative diseases, and the authors concluded that “newly developed significant TR is a rare event after successful repair of degenerative MR.”Based on our findings, postoperative functional TR is uncommon after MV repair for degenerative diseases, and we have identified factors associated with its development except for the role of the diameter of the TV annulus. Approximately 22% of our patients had advanced myxomatous degeneration with voluminous MV leaflets and large annulus, and these patients frequently also have voluminous tricuspid leaflets. Thus, a single measurement of 40 mm as the cut-off point of dilated TV annulus is probably inappropriate. It may be too large for patients with small TV leaflets such as in patients with MR due to fibroelastic deficiency and too small for patients with MR due to advanced myxomatous degeneration. We believe that further studies are needed to define the implications of TV annular dilatation.
Finally, TA is associated with recurrent TR in a relatively high proportion of patients as documented in this study, but we and others have shown that the type of TA is an important determinant of recurrent TR, and rigid rings and bands provide more durable repair than suture annuloplasty.
23
, 24
Limitations of the Study
As with most retrospective studies, this one has several limitations. Patients operated on during the first decade of this experience did not have detailed echocardiographic reports on the TV. Echocardiographic studies were performed in multiple laboratories, and we used the written report to enter postoperative MR and TR grades into our database. Together, these 2 points raise the possibility that we have underestimated the prevalence of TR both preoperatively and in the postoperative period in this study. In addition, although this study provides a unique long-term perspective on outcomes after MV repair, an era effect cannot be excluded through which risk factors once considered important no longer are and newly identified risk factors might have an effect that was not taken in account in the current study. We had no data on heart rhythm before the day of surgery or postoperatively until in the latter parts of the follow-up, preventing us from better defining the role of postoperative AF as well as the maze procedure in the development of TR. Finally, we must acknowledge that the patient population examined represents the practice of one surgeon. Our statistical analyses attempted to risk adjust within the unique context of our patient population, but we make no claims regarding the generalizability of those risk models to the wider patient population either clinically or statistically.
Conclusions
Patients with MR due to degenerative diseases develop TR because of older age, chronic AF, advanced functional class, impaired left ventricular function, congenital heart septal defects, and female sex. These factors adversely affect long-term survival, and TA does not seem to restore life span to the level of patients without TR. Patients with moderate and severe TR at the time of MV repair probably should have concomitant TA to reduce the probability of developing TR. New TR after MV repair is uncommon during the first 15 years of follow-up, but when it happens it is associated with poor prognosis largely because of the factors associated with it. The findings of this study are compelling reasons to recommend MV repair early on the course of severe MR due to degenerative diseases.
Conflict of Interest Statement
Authors have nothing to disclose with regard to commercial support.
Appendix
Table E1Variables examined for entry into risk models
| Variable examined for entry into risk models | Percent imputed |
|---|---|
| Year of operation | 0.0% |
| Age at operation (y) | 0.0% |
| Sex | 0.0% |
| Previous cardiac intervention | 0.0% |
| Elective (vs urgent) procedure | 2.5% |
| Preoperative angina/myocardial infarction | 0.0% |
| Preoperative NYHA class | 0.0% |
| Preoperative LV grade (<20%, 20%-39%, 40%-59%, ≥60%) | 0.1% |
| Preoperative atrial fibrillation | 0.1% |
| Preoperative congestive heart failure | 0.0% |
| Preoperative hypertension | 0.0% |
| Preoperative hyperlipidemia | 0.0% |
| Preoperative stroke/transient ischemic attack | 0.0% |
| Smoking history | 0.1% |
| Family history of coronary artery disease | 0.1% |
| Preoperative endocarditis | 0.0% |
| Congenital heart disease | 0.0% |
| Coronary artery disease | 0.0% |
| Mitral valve prolapse | 6.9% |
| Preoperative TR grade | 2.7% |
| Size of the mitral valve annulus (mm) | 10.8% |
| Type of mitral valve prosthesis | 0.0% |
| Concomitant tricuspid annuloplasty | 0.0% |
| Concomitant maze procedure | 0.0% |
| Postoperative atrial fibrillation (up to hospital discharge) | 0.0% |
| Postoperative pacemaker insertion (up to hospital discharge) | 0.0% |
| Postoperative MR grade (<1 mo after procedure) | 3.6% |
| Postoperative TR grade (<1 mo after procedure) | 3.6% |
NYHA, New York Heart Association; LV, left ventricular; TR, tricuspid regurgitation; MR, mitral regurgitation.
∗ Not included in models for preoperative TR grade.
† Included only in models of long-term outcomes.
References
- Impact of concomitant tricuspid annuloplasty on tricuspid regurgitation, right ventricular function, and pulmonary artery hypertension after repair of mitral valve prolapse.J Am Coll Cardiol. 2015; 65: 1931-1938
- Is the air in Toronto, Rochester, and Cleveland different from that in London, Monaco, Leiden, Genk, Milan, and New York?.J Thorac Cardiovasc Surg. 2015; 150: 1040-1043
- When is tricuspid valve annuloplasty necessary during mitral valve surgery?.J Thorac Cardiovasc Surg. 2015; 150: 1043-1044
- Expert opinion is often hot air: it is not the air; it is the clinical evidence.J Thorac Cardiovasc Surg. 2015; 150: 1045-1046
- Late outcomes of mitral valve repair for mitral regurgitation due to degenerative disease.Circulation. 2013; 127: 1485-1492
- Simplici-T annuloplasty band for mitral valve repair for degenerative disease.Ann Thorac Surg. 2014; 98: 1551-1556
- American Society of Echocardiography. Recommendations for evaluation of the severity of native valvular regurgitation with two-dimensional and Doppler echocardiography.J Am Soc Echocardiogr. 2003; 16: 777-802
- Guidelines for reporting mortality and morbidity after cardiac valve interventions.J Thorac Cardiovasc Surg. 2008; 135: 732-738
- Fate of functional tricuspid regurgitation after mitral valve repair for degenerative mitral regurgitation.Circ J. 2013; 77: 2288-2294
- Functional tricuspid regurgitation at the time of mitral valve repair for degenerative leaflet prolapse: the case for a selective approach.J Thorac Cardiovasc Surg. 2011; 142: 608-613
- Influence of mitral valve repair versus replacement on the development of late functional tricuspid regurgitation.J Thorac Cardiovasc Surg. 2014; 148: 1957-1962
- Clinical outcome of isolated tricuspid regurgitation.JACC Cardiovasc Imaging. 2014; 7: 1185-1194
- Significant lead-induced tricuspid regurgitation is associated with poor prognosis at long-term follow-up.Heart. 2014; 100: 960-968
- Do transvalvular pacemaker leads influence functional outcome after tricuspid ring annuloplasty?.Eur J Cardiothorac Surg. 2015; 48: 363-369
- Impact of the maze procedure and postoperative atrial fibrillation on progression of functional tricuspid regurgitation in patients undergoing degenerative mitral repair.Eur J Cardiothorac Surg. 2013; 43: 520-525
- Secondary tricuspid regurgitation or dilatation: which should be the criteria for surgical repair?.Ann Thorac Surg. 2005; 79: 127-132
- Pathogenic structural heart changes in early tricuspid regurgitation.J Thorac Cardiovasc Surg. 2015; 150: 323-330
- Guidelines on the management of valvular heart disease (version 2012).Eur J Cardiothorac Surg. 2012; 42: S1-S44
- Tricuspid annuloplasty prevents right ventricular dilatation and progression of tricuspid regurgitation in patients with tricuspid annular dilatation undergoing mitral valve repair.J Thorac Cardiovasc Surg. 2011; 141: 1431-1439
- The Society of Thoracic Surgeons Mitral Repair/Replacement Composite Score: a report of The Society of Thoracic Surgeons Quality Measurement Task Force.Ann Thorac Surg. 2016; 101: 2265-2271
- Accuracy and interobserver concordance of echocardiographic assessment of right ventricular size and systolic function: a quality control exercise.J Am Soc Echocardiogr. 2012; 25: 709-713
- Tricuspid annular size and regurgitation progression after surgical repair for degenerative mitral regurgitation.Am J Cardiol. 2016; 118: 424-431
- Tricuspid valve repair with an annuloplasty ring results in improved long-term outcomes.Circulation. 2006; 114: I577-I581
- Surgical management of secondary tricuspid valve regurgitation: annulus, commissure, or leaflet procedure?.J Thorac Cardiovasc Surg. 2010; 139: 1473-1482
Article Info
Publication History
Published online: February 07, 2017
Accepted:
December 16,
2016
Received in revised form:
November 30,
2016
Received:
May 18,
2016
Footnotes
This project was funded by the Academic Enrichment Fund of the Division of Cardiovascular Surgery of Toronto General Hospital and partially funded by a generous grant from Anthony and Miranda Wong.
Identification
Copyright
© 2017 Published by Elsevier Inc. on behalf of The American Association for Thoracic Surgery
User License
Elsevier user license | How you can reuse 
Elsevier's open access license policy
Elsevier user license
Permitted
For non-commercial purposes:
- Read, print & download
- Text & data mine
- Translate the article
Not Permitted
- Reuse portions or extracts from the article in other works
- Redistribute or republish the final article
- Sell or re-use for commercial purposes
Elsevier's open access license policy
30164-2&id=fx1.jpg){kind=link}
30164-2&id=gr1.jpg){kind=link}
30164-2&id=gr2.jpg){kind=link}
30164-2&id=gr3.jpg){kind=link}
30164-2&id=gr4.jpg){kind=link}
30164-2&id=fx2.jpg){kind=link}