|
|
||||||||
Eur J Cardiothorac Surg 2003;23:86-92
© 2003 Elsevier Science NL
Hadassah University Hospital, Cardiac Surgery Research Center, P.O. Box 12000, Jerusalem 91120, Israel
Received 25 June 2002; received in revised form 27 August 2002; accepted 1 October 2002.
* Corresponding author. Tel.: +972-2-6776-790; fax: +972-2-6780-927
e-mail: jbborman{at}netvision.net.il
| Abstract |
|---|
|
|
|---|
Key Words: Heart valve replacement Bicarbon prosthesis Bileaflet hinged mechanical device Results
| 1. Introduction |
|---|
|
|
|---|
The present report evaluates pooled results in a large series of 1875 patients who underwent Sorin BicarbonTM (SB) valve implantation in 14 centers over a 10 year period.
| 2. Patients and methods |
|---|
|
|
|---|
All patients met the accepted criteria for mechanical valve replacement [10]. Those who had implantations of SB plus another valve type were excluded from the study. The etiology of the valve disease is shown in Table 1. Rheumatic heart disease was the most common causative factor followed by degenerative pathology.
|
|
Outcomes were reported according to established international guidelines [11]. Events occurring within 30 days of surgery were designated early. All subsequent complications were classified as late.
Clinical evaluation was carried out yearly after the first year. Surgery closure date for this study was December 31, 1996. Data collection closing date was December 31, 2000. Relevant information was gathered by direct physical examination in the vast majority of cases. When this was not possible telephone communication or mailed questionnaires were substituted. Completeness of follow-up is 98.9%. Cumulative follow-up is 8280 years overall: 4462 AVR, 2893 MVR and 925 double valve replacement (DVR).
2.1. Statistical analysis
Estimated survival and freedom from event rates were calculated by the KaplanMeyer method (confidence limits 95%) [12].
Linearized complication rates were calculated as the number of events per 100 patient-years (%/patient-year). The
2-test was used for statistical comparison of NYHA Class groups. All statistical computations were performed with the PATS (Dendrite Clinical System, Portland, OR) software.
As actual complication rates have been reported to differ somewhat from actuarial outcomes, we calculated both. The actual complication rates were calculated according to a recent publication by Grunkemeier et al. [13].
| 3. Results |
|---|
|
|
|---|
3.1. Survival
Estimated survival curves by implant site including early deaths are shown in Fig. 1
. Overall survival, including early deaths is AVR 71.8±4.5%, MVR 69.4±3.3% and DVR 81.4±3.1% at 8 years. Survival excluding early deaths is as follows: overall 75.8±2.9% (AVR 75.3±4.3%, MVR 75.1±3.5% and DVR 86.2±2.9% at 8 years). Valve related deaths overall totaled 55 patients. Actuarial overall freedom from valve related death at 8 years is 95.5±0.7%; MVR and AVR having remarkably similar results: AVR 95.2±1.3%, MVR 95.2±1.0% and DVR 96.9±1.4%. Causes of death are reported in Tables 3 and 4.
|
|
|
3.3. Valve thrombosis
Twenty-two events were recorded: three AVR, 15 MVR and four DVR. AVR has significantly better results as regards morbidity from this complication. Overall freedom from thrombosis is 98.5±0.3% at 8 years (99.6±0.3% and 97.2±0.7% for AVR and MVR respectively, and 97.1±0.6% for DVR). Linearized rates are reported in Table 5. Of the 22 instances of thrombosis, 11 occurred within 30 days. There was one fatal event, eight were managed by thrombolysis and 13 required reoperation (nine MVR, two AVR, two DVR).
|
3.5. Hemolysis
Significant hemolysis without paravalvular leak occurred in this series in only six instances: three AVR, one MVR and two DVR. DVR caused relatively increased hemolysis as compared to single valve replacement. Overall freedom from hemolysis was 98.3±0.9% at 8 years (AVR 98.1±1.7%, MVR 93.3±0.7%, and DVR 94.6±2.0%). All the patients survived and none required reoperation. Linearized rates are shown in Table 5.
3.6. Non-structural dysfunction
Twenty-six events were listed six AVR, 16 MVR and four DVR; all were paravalvular leaks (PPL). Twenty patients required operation and all survived. Overall freedom from non-structural dysfunction was 98.0±0.4% at 8 years (AVR 96.8±0.6%, MVR 96.8±0.8%, and DVR 97.5±1.2%). Linearized rates are shown in Table 5.
3.7. Anticoagulant related bleeding
Included are all hemorrhages leading to hospitalization. Sixty-six episodes were recorded: 41 AVR, 16 MVR and nine DVR. Overall freedom from major bleeding was 94.6±0.7% at 8 years (AVR 93.5±1.1%, MVR 96.9±0.8%, and DVR 92.6±2.7%). Mortality was 32%, mainly when the hemorrhage was intracerebral. Linearized rates are shown in Table 5.
3.8. Endocarditis
All prosthetic valve infections are included, whether primary or recurrent episodes (patients who had preoperative endocarditis). Thirty-one instances were reported 16 AVR, 14 MVR and one DVR. Fourteen did not respond to antibiotic therapy and were reoperated. Overall freedom from endocarditis is 97.8±0.4% at 8 years, (AVR 98.0±0.5%, MVR 97.2±0.7 and DVR 98.6±1.4%). Mortality was 29%.
3.9. Reoperation
Forty-nine patients were reoperated AVR 14, MVR 28 and DVR 7. Overall freedom from reoperation at 8 years is 96.6±0.5% and is shown in Fig. 2
. Severe PPL (20), infective endocarditis (14) and thrombosed valves (13) were the reasons for reoperation, two more cases were non-valve related. Most reoperations were carried out during the first 3 years after the initial surgery. There have been no reoperations to date, after the 6-year mark. Three patients in this group died during their hospitalization: one AVR, one MVR and one DVR; two died late. Mortality for reoperation is 10%. Linearized rates of all major complications are shown in Table 5.
|
| 4. Discussion |
|---|
|
|
|---|
Early (30 day) mortality in this series was 5%, which is essentially similar or somewhat less than in some other reports related to bileaflet valves [8,1416]. Our overall survival rates over 10 years (71.6±2.8%) were somewhat better than the figure of 63.3% reported for SJM valves [14], whereas Aagard [17] reported similar survival rates to ours for Carbo Medics (CM) prostheses. On the other hand Nakano et al. reported 1112 year actuarial survival of 81.7% for AVR, 87.1% for MVR and 82.6% in Japanese patients with SJM prostheses [18]. However, these results are not at all comparable as the incidence of concomitant (coronary bypass) surgery was only 1.4% compared to an incidence of concomitant surgery of 25% (coronary bypass 15%) in our series. Furthermore, the average age of Nakano's cohort was 48 compared to a median age of 60 in this series. These discrepancies demonstrate the difficulties and unreliability of comparing survival outcomes in different series, where patient characteristics often differ markedly. Overall actuarial survival excluding early deaths in our series is 75.8±2.9% and actuarial freedom from valve related deaths is satisfactorily high: 95.5±0.7% at 8 years. Actuarial freedom from valve-related mortality over 9 years for CM valves was 91.0±2.9% for AVR and 93.5±1.8% for MVR [16]; for SJM valves the respective figures at 9 years are 79±11 and 91±3% [14]. Aoyagi reported freedom from valve related deaths at 10 years with SJM valves: AVR 94±2.0%, MVR 89±2.0% [15]. SJM results reported by Nakano in his young, free-of-coronary-disease Japanese patients were again better than reports from Western countries: AVR 97.1±1.1% and MVR 97.6±6.7% [18]. It is recognized and well-documented that concomitant coronary artery bypass surgery and older age are important risk factors effecting long term survival [19,20]. Actuarial survival figures at 8 years in our experience with the SB are relatively high, comparing favorably with the SJM and CM devices.
A major challenge related to mechanical valve manufacture concerns minimization or elimination of valve thrombosis and embolization. All mechanical valves are potentially thrombogenic; consequently well-controlled, effective anticoagulant therapy is absolutely mandatory. With this in mind, the SB valve was designed to maximally improve flow and washout characteristics. In our experience there were 22 thrombotic events, mostly related to valves in the mitral area. Eight patients responded to lytic agents; 13 required reoperation, one (MVR) died. Unfortunately, this patient was not referred for either lytic or surgical management.
Aoyagi reported four instances of valve thrombosis in a cohort of 908 SJM valve replacements in a younger age group (average 51.6 years) [15]. Dalrymple et al. [20] reported six instances of valve thrombosis in a sample of 1503 CM patients, but the follow up period was only 5 years. Jamieson comparing CM and SJM prostheses reported that patients aged >65 were more prevalent to major thromboembolism and hemorrhage, while type of valve implanted was insignificant [16]. The linearized rate for this complication in our series is 0.27%/patient-years overall: 0.07 AVR, 0.52 MVR and 0.43 DVR (Table 5). These results are similar (0.30.33%/patient-years) to those reported for SJM valves by Czer [14] and Baudet [21] and for CM valves by Jamieson [16]. Li referring to AVR published 99% freedom from thrombosis at 5 years [22].
Our database registered 95 embolic events, both minor and major. Implant position was not a significant factor. There were 27 reports of cerebral embolus causing major brain damage and 13 of these proved fatal. Transient cerebral ischemia occurred in 51 instances. Other emboli numbered 17, with three deaths. Overall freedom from all embolic events was 90.7±1.2% at 8 years. Our linearized rate for all embolic events is 1.15±0.21%/patient-years. Linearized rates from 0.8%/patient-years to 2%/patient-years have been reported for embolic events associated to other bileaflet valves [1421]. Thromboembolism occurrence (1.32%/patient-years) in our series is in the lower range of published data [1424]. However, any comparison between different reports may be unreliable in light of the large variability in results reported for thromboembolism related to prosthetic valves [1323]. Follow-up extent and reporting methods, adequacy of anticoagulation management, patient risk factors may contribute to the wide range of reported events.
Anticoagulant related bleeding is another aspect of the problem mentioned above the minimization of thromboembolic complications/bleeding related to implantation of mechanical intracardiac prostheses. The challenge at present is to manage each patient to achieve his/her ideal anticoagulation point where no thrombosis occurs and thromboembolism is kept to the absolute minimum for the lowest level of bleeding complications [23]. We report on all hemorrhages necessitating hospitalization, 66 in number. One third (22) involved intracerebral bleeding leading to death in 19 of these. Eight year freedom from bleeding is 93.5±1.1% for AVR, 96.9±0.8% for MVR and 92.6±2.7% for DVR. The linearized rate for anticoagulant related bleeding for the SB valve is 0.80±0.18%/patient-years. This incidence is similar to that reported by Aoyagi [15] and others [12,21]. However, bleeding rates
2%/patient-years have been observed in patients with SJM [16,19] and CM [16,17,20] valves. It is abundantly clear that intensified, continuous long-term efforts are absolutely essential in every patient with a mechanical valve in order to achieve Akins ideal anticoagulation point mentioned above [23].
All mechanical valves, being foreign bodies in the bloodstream with unnatural flow patterns, cause damage to blood elements viz. erythrocytes and platelets. Compared with other valves the SB valve produced lower damage indices [25]. Decompensated chronic hemolytic anemia is rare and usually associated with paravalvular leaks [9]. However, hemolysis without evidence of leakage has been reported with SJM valve [15,18]. Such complication in our SB collaborative study appeared in 6 cases. Overall linearized rate of hemolysis is 0.07%/patient-years which compares favorably with reported (0.2%/patient-years) for SJM valves [1517].
Non-structural dysfunction is unrelated to intrinsic valve dysfunction and for practical purposes is due to periprosthetic leakage (PPL). Such leaks are related to human factors such as heavily calcified, distorted native valve annuli and/or deficiencies in suturing technique [9]. Our linearized incidence of paravalvular leak is 0.31%/patient-years. All patients who required reoperation survived.
Endocarditis was diagnosed in 31 cases and is a constant danger, unrelated to length of time after implant. No specific site of implant was identified to be more susceptible to infection. Nine infections proved to be fatal and mortality was high when reoperation was performed, which was necessary in 14 patients. The overall occurrence of linearized rates 0.37±0.13%/patient-years is similar to that presented for SJM and CM valves [21,22].
Nearly all reoperations in this large combined series were performed within the first few years after the initial surgery. Severe PPL, thrombosed valve and endocarditis were the main reasons for reoperation. The mitral valve was reoperated most frequently and mortality was highest when reoperating for infective endocarditis. Overall occurrence of reoperation was 0.59±0.16%/patient-years and is similar to that reported for SJM [18] and CM [20] valves.
Functional improvement after surgery was consistent: overall 78% improved, 17% remained stable and 5% deteriorated.
For illustrative purpose non-fatal events were also analyzed using actual statistical techniques with regard to the following events: endocarditis, non-structural dysfunction, anti-coagulated bleeding, thrombosis, embolism and hemolysis (Table 5). Our findings confirm the assertion of a somewhat lower incidence of events using actual (cumulative) analyzes. As an example Fig. 3 relates to embolic events and clearly demonstrates the small but increasing divergence with the passage of time to the benefit of the actual method.
|
| 5. Conclusions |
|---|
|
|
|---|
| Acknowledgments |
|---|
| References |
|---|
|
|
|---|
This article has been cited by other articles:
![]() |
D. Baykut, L. Grize, C. Schindler, A. S. Keil, F. Bernet, and H.-R. Zerkowski Eleven-year single-center experience with the ATS Open Pivot Bileaflet heart valve. Ann. Thorac. Surg., September 1, 2006; 82(3): 847 - 852. [Abstract] [Full Text] [PDF] |
||||
![]() |
M. Josa, M. Castella, C. Pare, J. L. Bedini, R. Cartana, C. A. Mestres, J. L. Pomar, and J. Mulet Hemolysis in Mechanical Bileaflet Prostheses: Experience With the Bicarbon Valve Ann. Thorac. Surg., April 1, 2006; 81(4): 1291 - 1296. [Abstract] [Full Text] [PDF] |
||||
![]() |
N. Vitale, G. Cappabianca, G. Visicchio, C. Fondacone, V. Paradiso, G. Mannatrizio, and L. d. L. T. Schinosa Midterm evaluation of the Sorin Bicarbon heart valve prosthesis: single-center experience Ann. Thorac. Surg., February 1, 2004; 77(2): 527 - 531. [Abstract] [Full Text] [PDF] |
||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| ANN THORAC SURG | ASIAN CARDIOVASC THORAC ANN | EUR J CARDIOTHORAC SURG |
| J THORAC CARDIOVASC SURG | ICVTS | ALL CTSNet JOURNALS |