| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Ann Thorac Surg 2006;81:1317-1324
© 2006 The Society of Thoracic Surgeons
Department of Thoracic and Cardiovascular Surgery, Cardiovascular Center, Severance Hospital, Yonsei University College of Medicine, Seoul, South Korea
Accepted for publication November 3, 2005.
* Address correspondence to Dr Chang, Division of Cardiovascular Surgery, Yonsei Cardiovascular Center, Yonsei University College of Medicine, 134 Shinchon-dong, Seodaemun-gu, Seoul, South Korea 120-752 (Email: bcchang{at}yumc.yonsei.ac.kr).
Presented at the Forty-first Annual Meeting of The Society of Thoracic Surgeons, Tampa, FL, Jan 24–26, 2005.
 |
Abstract |
|---|
 Top Abstract IntroductionPatients and MethodsResultsCommentDiscussionReferences |
|---|
METHODS: Between 1978 and 2003, we performed 138 TVR (35 bioprosthetic, 103 mechanical) in 125 patients (50 men, 75 women), with a mean age of 43.7 ± 16.6 years. The diseases that required TVR were rheumatic (94), prosthetic valve failure (14), congenital (14), infective endocarditis(5), isolated tricuspid regurgitation (4), and miscellaneous conditions (7). The operations included the following: isolated TVR (41), double valve replacement (58), and triple valve replacement (39). The follow-up rate was 98.3%, and cumulative follow-up was 828.5 patient-years.
RESULTS: There were 22 in-hospital deaths (17.6%) and 13 (10.4%) late deaths. Fourteen patients required additional operations. There were 33 postoperative valve-related events including 11 thromboembolisms and 3 bleeding episodes. Kaplan-Meier survival for the entire group at 15 years was 73.8 ± 8.5% (bioprosthetic: 70.2 ± 10.4%, mechanical: 66.0 ± 19.4%). At 15 years, freedom from reoperation was 66.3 ± 9.4% (bioprosthetic: 55.1 ± 13.8%, mechanical: 86.0 ± 6.2%) and freedom from valve-related events was 49.9 ± 8.0%. The linearized incidence of valve thrombosis was 1.28%/patient-year (bioprosthetic: 0, mechanical: 1.92), anticoagulation-related bleeding was 0.37%/patient-year (mechanical: 0.54), reoperation was 1.71%/patient-year (bioprosthetic: 2.68, mechanical: 1.25), and valve-related events were 4.33%/patient-year (bioprosthetic: 3.83, mechanical: 4.6).
CONCLUSIONS: Both bioprosthetic and mechanical valves revealed similar long-term outcomes. However, findings suggest that greater care is needed to prevent valve thrombosis in mechanical valves in the early postoperative period, and there is a greater chance for reoperation in bioprosthetic valves.
|
Introduction |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsCommentDiscussionReferences |
|---|
|
Patients and Methods |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsCommentDiscussionReferences |
|---|
|
|
Anticoagulation
Heparin infusion was started at the first or second postoperative day, unless there were clinical contraindications, and sodium warfarin was started on the second postoperative day. The prothrombin time was maintained at an international normalized ratio (INR) of 2.5–3.5 for a mechanical valve, and an INR of 1.5–2.5 for a bioprosthetic valve.
Patient Follow-Up
One hundred sixteen patients were followed up periodically by the cardiologists, surgeons, or referring physicians. Follow-up was accomplished by patient visit, by mailing, or by telephone with the patient or the patient's family, with a follow-up rate of 98.3%. The mean follow-up period was 7.1 ± 5.0 years (9.7 ± 6.7 years for bioprostheses and 6.4 ± 4.1 years for mechanical), and the cumulative follow-up was 828.5 patient-year (261.3 for bioprosthesis and 567.2 patient-year for mechanical valve).
Statistical Analysis
Valve-related complications were defined in accordance with the guidelines established by the American Association for Thoracic Surgery and the Society of Thoracic Surgeons [8]. Variables of the two groups were compared using an unpaired t test. Estimates of survival and event-free survival, and mortality were calculated using the Kaplan-Meier method, and are reported with 95% confidence limits. Late valve-related events were defined as thromboembolism, valve thrombosis, anticoagulation-related bleeding, prosthetic valve endocarditis, paravalvular leakage in the absence of infection, and pannus formation. Valve-related complications were expressed in linearized form (%/patient-year). Estimates were reported with their standard errors. Comparisons of the curves were established using the Wilcoxon test for early and midterm results, and the log-rank test for long-term results. To analyze the risk factors for early and late mortality we reviewed preoperative variables, which are listed in Table 1. Risk factor analysis was performed by 
2 and Fisher's exact testing, and an unpaired t test was used for univariate analysis. Multivariate analysis was performed using a logistic regression method. Results were calculated using a statistical software package (SAS Version 8.1; SAS Institute, Cary, NC). A p value less than 0.05 was considered to be significant for all analyses.
|
Results |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsCommentDiscussionReferences |
|---|
|
Late Death and Patient Survival
Among the 116 patients discharged from the hospital, 13 patients (6 in bioprosthetic [22.2%], 7 in mechanical [7.9%]) died during the follow-up period (late mortality; 11.4%). Nine of the 13 patients had received a previous valve replacement more than once. Five patients received single valve replacement (4 bioprostheses and one mechanical), 6 patients received double valve replacement (2 bioprostheses and 4 mechanical), and two patients received triple valve replacement (2 mechanical). The most common cause of late death was unknown (n = 7) and postmortem examinations were not performed in any cases. However, most of the patients had suffered from heart failure previously, and the causes of these deaths were thought to be underlying heart failure. Other causes of late mortality are shown in Table 2. The Kaplan-Meier survival rates were 94.2 ± 0.6% at five years, 84.3 ± 0.7% at 10 years, and 73.8 ± 1.1% at 16 years (Fig 2). There was no statistical difference in survival rates between the two groups (Wilcoxon, p = 0.1759; log-rank, p = 0.2796). The linearized incidence of late mortality for all patients was 1.57%/patient-year (2.30%/patient-year for bioprostheses vs 1.23%/patient-year for mechanical valves). Univariate analysis identified peripheral edema as a risk factor, while multivariate logistic analysis identified preoperative ascites and peripheral edema as risk factors for late death (Table 3).
|
|
|
Prosthetic Valve Failure
During the follow-up periods, prosthetic valve failure occurred in 11 patients, including seven structural failures in the bioprosthetic and four pannus formations in the mechanical groups. Pannus formation occurred in patients once each with the St Jude and ATS valves, and twice with DuroMedics valves. All structural bioprostheses valve failures were due to valve cusp tearing or severe leaflet calcification between 83.9 and 196.5 months after TVR. In all cases re-replacements were performed using a mechanical valve. The linearized incidence of structural valve failure was 0.84%/patient-year for all patients and 2.68%/patient-year for the bioprostheses group. The linearized incidence of nonstructural valve failure was 0.48%/patient-year for all patients and 0.71%/patient-year for the mechanical group.
Reoperation
Reoperation was required in 14 patients. The causes of reoperation were structural valve failure in the bioprostheses group (7), and pannus formation (4), valve thrombosis (1), re-replacement of tissue valve during mitral valve replacement using a mechanical valve (1), and paravalvular leakage (1) in the mechanical group. Among 14 reoperations, 13 patients underwent tricuspid valve re-replacement and one patient underwent repair for paravalvular leakage. The Kaplan-Meier analysis for freedom from reoperation is shown in Figure 4. In the bioprosthetic group survival without reoperation declined 10 years after operation. The linearized incidence of reoperation for all patients was 1.71%/patient-year (2.68%/patient-year in bioprosthetic, 1.25%/patient-year in mechanical).
|
Freedom From Valve-Related Events
The rate of event-free survival is shown in Figure 5. There was no statistical difference between the two groups. The linearized incidence of valve-related events for all patients was 4.33%/patient-year (3.83%/patient-year in bioprostheses, 4.6%/patient-year in mechanical). There was no statistical difference between the two groups.
|
|
Comment |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsCommentDiscussionReferences |
|---|
Severe TR related to left side heart valve disease may be one of the important manifestations of heart failure and early surgical correction should be considered. Tricuspid valve surgery concomitant with left side heart valve surgery is associated with a high operative mortality, especially in patients requiring TVR. Most of these patients received one or more previous procedures, including tricuspid valve repair, and were in advanced stages of functional deterioration [1]. For the past several decades, results of heart valve surgery have improved significantly with the development of myocardial protection and surgical techniques. However, surgical results of TVR are still poor and need to be improved. The surgical mortality rate of 15.9% in this series is very high compared with mitral or aortic valve replacement but is comparable with previously reported TVR series, which range from 7.7% to 27% [2–15]. The high operative mortality of TVR appears to have several important factors. As we experienced previously in patients with mitral regurgitation, patients can tolerate severe TR for a long time as well. However, when the symptoms of right heart failure, such as peripheral edema and ascites occur, we can anticipate that their conditions may rapidly deteriorate, thus decreasing the likelihood of a positive surgical result. Recently our policy has been early surgical correction for severe TR in order to prevent progressive right heart failure.
Our results show that surgical mortality for bioprostheses was higher than that of mechanical prostheses, but without statistical significance (p = 0.1985). This result is comparable with previous reports [6, 7, 12]. The most common cause of early death in this series was persistent underlying heart failure (54.5% of operative deaths). Although recent advances in myocardial protection during surgery might reduce the incidence of myocardial failure during the early postoperative period, myocardial dysfunction before surgery seems to be a major cause of death. Operation before the development of ventricular failure is thought to be important for reducing early death after TVR.
The long-term survival after TVR with mechanical valves has been reported to be similar to TVR with tissue valves [9, 10]. In this series there was also no statistical difference between two types of valve prostheses in terms of long-term survival. The major cause of late death was preexisting heart failure. If these patients had been treated before developing heart failure, a better long-term survival may have been expected.
The late mortality of 11.4% is comparable with or lower than other reports, which range from 9.2% to 57% [2, 7, 12, 16]. Some authors reported that there was no late mortality during their follow-up period [5, 12]. In these 13 patients, nine underwent two or more valve operations. Among the 13 deaths, two patients died of heart failure and seven patients died of unknown causes. But a significant number of the patients with unknown causes of death were in poor clinical condition before death. Thus the unknown deaths are thought to be related to the progression of underlying heart failure. Therefore, early operation before development of heart failure seems to be very important. Regarding the type of prosthesis, there was no statistical difference between the two groups in valve-related mortality (p = 0.4029). However, there was a tendency to develop more valve-related mortality in the bioprosthetic group (11.1% in bioprosthetic vs 4.5% in mechanical valve group).
Valve thrombosis and pannus formation have been important valve-related complications after TVR with mechanical valves. The higher incidence of tricuspid valve thrombosis in previously designed tilting disc mechanical valves has been confirmed by several authors [17–21]. Recently, Kawano and colleagues [5] reported six cases of valve thrombosis among 19 patients who had received TVR with the St. Jude Medical valve (linearized incidence of 2.9%/patient-year). Two of these patients did not take warfarin sodium regularly. Dalrymple-Hay and colleagues [6] also reported a high prevalence of valve thrombosis in TVR with mechanical valves. However, Nakano and colleagues [3] reported only one valve thrombosis with 96.9% freedom from valve thrombosis 14 years after surgery. Furthermore, Singh and colleagues [4] reported no occurrence of valve thrombosis. In this report 102 bileaflet pyrolytic carbon valves were used, with one exception of tilting disk. There were 10 episodes of valve thrombosis between one and 46 months after TVR. At 48 months after TVR, there was no further valve thrombosis. Among 10 valve thromboses, three cases occurred in the patients with the DuroMedics valve (3 in 8 patients), which may be related to structural problems of this valve or surgical techniques. The linearized incidence of valve thrombosis in the mechanical valve group was 1.92%/patient-year, which is higher than previously reported results of other valve positions (MVR: 0.54%/patient-year and AVR: 0.33%/patient-year) [22]. As Kawano and colleagues [5] already described, there may be a higher incidence of prosthetic valve thrombosis with the low-pressure chamber at the tricuspid position. Anticoagulation measures should be strictly regulated even with low thrombogenic bileaflet pyrolytic carbon prostheses.
Valve thromboses and thromboembolic complications are affected by several factors, such as ventricular function, rhythm, valve position, prosthetic valve type, coagulation status (protein C, S, AT III, etc, level), and patient compliance, which may be the most important factor. As we have previously mentioned in another study [22], tight control of individualized anticoagulation level and regular education of the patients are the most important measures to prevent thrombotic complications. Further studies for the development of an ideal prosthesis to prevent valve thrombosis and pannus formation are necessary.
Reoperation for tricuspid position is an important consideration during prosthetic valve selection. In this study, there was no statistical difference between the two types of prostheses 16 years after operation in terms of reoperation. Reoperations took place at different times depending on prosthesis type, with reoperations in the mechanical valve group occurring within seven years, while reoperations in the bioprosthetic valve group occurred beyond 12 years, with one exception. Rizzoli and colleagues [9] reported that the risk of reoperation was 4.7 %/patient-year for bioprostheses and 2.2 %/patient-year for mechanical prostheses. They also reported that bioprosthetic valve degeneration increased at a steeper rate after seven years. Dalrymple-Hay and colleagues [6] insisted that the first 10 years after TVR are crucial in the decision making process for selecting an optimal implant, and that the risk of reoperation due to structural degeneration of a bioprosthetic valve is significantly lower than the risk of death during this period. In this study, freedom from reoperation using the mechanical prostheses 16 years after the initial operation (86.0 ± 1.4%) was comparable to other reports [5–7]. Although there was no statistical difference, the linearized incidence of reoperation was somewhat higher in the bioprosthetic valve group (2.68%/patient-year) than in the mechanical valve group (1.25%/patient-year).
Still, controversy exists as to the most suitable prosthesis for the tricuspid position. Some authors have reported good results with bioprostheses [5–7, 14], while others have shown good results with mechanical valves [2, 12, 23]. Early operation must be considered in patients with bioprostheses before significant degeneration of the bioprostheses and ventricular dysfunction develops. We also must consider the balance between the risk of valve-related complications when using the mechanical valve and the risk of reoperations when using the bioprosthetic valve.
Tricuspid valve replacement still carries a high operative mortality and poor long-term survival. The poor results are influenced by preoperative and postoperative patient conditions especially in patients with preexisting right heart failure. Correcting tricuspid regurgitation before irreversible heart failure occurs is thought to be important. Both bioprosthetic and mechanical valves demonstrated similar long-term clinical results in the tricuspid position. However, findings suggest that greater care is needed to prevent valve thrombosis in pyrolytic carbon-coated bileaflet mechanical valves, and greater concern is warranted for reoperation in bioprosthetic valves.
|
Discussion |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsCommentDiscussionReferences |
|---|
Also, it looked like at 16 years there was a difference, though not statistically significant, in reoperation-free survival between tissue and mechanical valves, which we probably would expect, and I wonder if you would comment on that? Finally, at the present time, what is your clinical practice with implantation of valves into the tricuspid position?
DR CHANG: Thank you. For the first question and comment, thank you very much, the tissue valve has small numbers, only 35 cases; however, we followed them up more than 20 years, and 10 years after most of the patients required reoperation. Although the statistical power is low, the requirement of reoperation is necessary in most of the patients.
And for the second question, the long-term results after five years look better with mechanical valves, and also recently a high profile housing mechanical valve has come out and we may assume that it will prevent pannus formation for patients requiring tricuspid valve. Thus, if the patient is young, we would choose a mechanical valve instead of a tissue valve for prevention of reoperation.
DR JONATHAN HAMMOND (Hartford, CT): Probably most of us either have directly dealt with or have a partner who has dealt with an acutely thrombosed mechanical valve in the tricuspid position, and it is not a pretty sight. Do you think in the current era of the bovine pericardial valve that some of the durability problems might be solved and we would still feel comfortable with putting bioprosthetic valves in the tricuspid position?
DR CHANG: Thank you. Among 35 cases, half of them received a valve replacement before 1990. So most of them had a first generation tissue valve. We believe that the pericardial tissue may be more durable. However, recently we found that pericardial valves have some problems for durability in several patients, especially in young patients. So we have to follow them up more I think.
|
References |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsCommentDiscussionReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  R. Dierckx, S. E. Haine, C. J. Vrints, and B. P. Paelinck Young Adult With Congenital Heart Disease Presenting With Anasarca Circulation, September 16, 2008; 118(12): 1304 - 1305. [Full Text] [PDF]
|
|
|
|
 |
|
 K. Sughimoto, K. Nakano, A. Gomi, H. Nakatani, Y. Nakamura, and A. Sato Hancock Valve Deterioration in Tricuspid Position for Ebstein's Anomaly Asian Cardiovasc Thorac Ann, April 1, 2008; 16(2): 157 - 158. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 I. Rafiq, L. J. Freeman, and S. Wood Importance of Long-Term Anticoagulation in Both Prosthetic and Biological Tricuspid Valve Replacements Ann. Thorac. Surg., October 1, 2007; 84(4): 1424 - 1424. [Full Text] [PDF]
|
|
|
|
|
|
S. Lee and B.-C. Chang Reply Ann. Thorac. Surg., October 1, 2007; 84(4): 1425 - 1425. [Full Text] [PDF]
|
|
|
|
 |
|
 Z. H. Iscan, K. M. Vural, I. Bahar, L. Mavioglu, and A. Saritas What to expect after tricuspid valve replacement? Long-term results Eur. J. Cardiothorac. Surg., August 1, 2007; 32(2): 296 - 300. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. H. Rahimtoola The Year in Valvular Heart Disease J. Am. Coll. Cardiol., January 23, 2007; 49(3): 361 - 374. [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 |
