| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
J Thorac Cardiovasc Surg 2008;135:878-884
© 2008 The American Association for Thoracic Surgery
Surgery for Acquired Cardiovascular Disease |
a Division of Cardiovascular Surgery, Mayo Clinic, Rochester, Minn
b Division of Biostatistics, Mayo Clinic, Rochester, Minn
Received for publication June 23, 2007; revisions received October 15, 2007; accepted for publication October 26, 2007. * Address for reprints: Hartzell V. Schaff, MD, Mayo Clinic, 200 1st St SW, Rochester, MN 55905. (Email: schaff{at}mayo.edu).
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionConclusionsReferences |
|---|
Methods: In this retrospective study, patients received either St Jude bileaflet valves or Carpentier–Edwards bioprostheses. Operations were performed between January 1991 and December 2000, and groups were matched one-to-one according to age, sex, need for coronary artery bypass grafting, and valve size.
Results: Four hundred forty patients were matched, and follow-up was 92% complete, with median durations of 9.1 years for patients who received mechanical valves and 6.2 years for patients who received bioprostheses. The 5- and 10-year unadjusted survivals were 87% and 68% for mechanical valves and 72% and 50% for bioprostheses, respectively (P < .01). Freedoms from reoperation at 10 years were 98% for mechanical valves and 91% for bioprostheses (P = .06). Rates of late stroke or other embolic events and of endocarditis were similar between groups. Hemorrhagic complications necessitating hospitalization occurred in 15% of patients with mechanical valves and 7% of patients with bioprostheses (P = .01). Notably, 19% of patients with bioprostheses were receiving warfarin sodium at last follow-up. After adjustment for unmatched variables, including diabetes, renal failure, lung disease, New York Heart Association functional class, ejection fraction, and stroke, the use of a mechanical valve was protective against late mortality (hazard ratio 0.46, P < .01).
Conclusion: In this study, patients aged 50 to 70 years who underwent aortic valve replacement with mechanical valves had a survival advantage relative to matched patients who received bioprostheses. These findings question recommendations of bioprostheses for younger patients and suggest that a randomized trial may be warranted.
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionConclusionsReferences |
|---|
recommend mechanical prostheses for aortic valve replacement (AVR) in patients younger than 65 years to avoid the risks of reoperation. Many surgeons believe that this age cutoff should be lowered because of the improved durability of bioprosthetic valves, which do not require anticoagulation with warfarin sodium.2-4 Evidence for this practice, however, is lacking. To the contrary, some believe that mechanical aortic valves should be offered to all patients up to 70 years of age because of the greater longevity of the general population and risks associated with reoperation.5 The selection of aortic valve prostheses for patients between 50 and 70 years old thus remains controversial. In this study, we investigated late mortality and clinical outcomes after insertion of a bioprosthetic or mechanical aortic valve in patients 50 to 70 years old.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionConclusionsReferences |
|---|
Patients were matched one-to-one according to age (±2 years), sex, presence of coronary artery bypass grafting, and valve size. A power calculation estimated that approximately 220 patients per group were required to have a minimum of 80% power to detect a 10% difference in mortality between the two groups with a 2-sided 
of .05. This power calculation assumed survival at 10 years after AVR to be approximately 65%. Patient characteristics were prospectively recorded in a clinical database at the time of surgery. These variables included demographic characteristics, risk factors, operative details, and early outcomes (within 30 days of surgery) according to the definitions in the Society of Thoracic Surgeons National Database.
Late outcomes (later than 30 days after surgery) were collected from the electronic medical record and from questionnaires mailed to patients by our survey research center. All positive answers were checked against the medical record whenever possible. Reports of additional cardiac operations were reviewed in detail. Each patient's vital status was checked through Accurint (LexisNexis, New York, NY) and the Social Security Death Index.
Statistics were performed with SAS statistical software (SAS Institute, Inc, Cary, NC). Results for continuous data are reported as mean ± SD or median with interquartile range as appropriate and for categoric data as the number and percentage of each category from the observed total. To compare categoric variables between valve groups 
2 tests (or Fisher exact tests for sparse data) were used, whereas t tests (or Wilcoxon rank sum tests for nonnormal data) were used to compare continuous parameters among preoperative, operative, early, and late complications. Kaplan–Meier actuarial survival methods were used to assess stroke, bleeding, late survival, and reoperation as a function of time since surgery. The groups were compared with a log-rank test. All estimates are provided with 95% confidence intervals (CIs). Cox proportional hazards regression was used to discriminate prognostic factors associated with time to death. Candidate variables considered for multivariate analysis were those detected by univariate models as having a significant association (P < .05) or suggestive trend toward association (P .05–.10) with mortality. To discriminate independent risk factors, multivariate modeling was performed with methods of stepwise selection, with valve group and candidate variables all competing for entry into a final model predicting late mortality.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionConclusionsReferences |
|---|
Figure 1). Follow-up was 92% complete, with mean durations of 8.6 years (median 9.1 years, interquartile range 5.99–11.95 years) for patients who received mechanical valves and 6.3 years (median 6.2 years, interquartile range 4.50–9.62 years) for patients who received bioprostheses.
|
Table 1. The patient groups were similar for all variables except age, which was 0.9 years higher in the bioprosthesis group (66.6 vs 65.7 years, P < .01). In addition, the mean cardiopulmonary bypass and crossclamp durations were statistically significantly higher, although not clinically significantly so, in the bioprosthetic group by 5.4 and 4.2 minutes, respectively (P = .03 for each; Table 1).
|
Table 2). There was a higher perioperative mortality in the bioprosthetic group (5.5% vs 1.8%, P = .04). In the mechanical valve group, the deaths were all caused by cardiac-related causes. In the bioprosthesis group, there were 7 cardiac-related deaths, 1 cardiac tamponade, 1 aortic dissection, 1 pulmonary failure, 1 liver failure, and 1 hemorrhagic stroke. There were also trends toward higher rates of reoperation for bleeding among patients receiving bioprosthetic valves (P = .08). There were no differences between the groups in early incidences of transient ischemic attack, stroke, reoperation for valvular problems, and readmission to hospital.
|
Table 3). The 10-year rate of freedom from stroke, however, was not statistically different between the bioprosthesis group (91.4%, CI 85.8%–95.8%) and the mechanical valve group (93.2%, CI 88.9%–97.3%; P = .07). There were also no differences in incidences of late transient ischemic attack and embolus between the mechanical valve and bioprosthesis groups (P = .72 and P > .99, respectively). Hemorrhagic complications necessitating hospitalization occurred in 15% of patients with mechanical valves and in only 7% of patients with bioprostheses (P = .01). Because of the shorter follow-up interval in the bioprosthesis group, however, we calculated the risks of bleeding at 10 years as 13.7% (CI 8.3%–19.2%) in the mechanical valve group and 6.4% (CI 1%–12.3%) in the bioprosthesis group (P = .06;
Figure 2).
|
|
Freedom from reoperation on the aortic valve was not significantly different between the mechanical valve and bioprosthesis groups at 10 years (97.5% vs 91%, P = .13;
Figure 3). Freedom from all late reoperations was not significantly different at 10 years: mechanical valve 96.7% (CI 83.9%–97.3%) and bioprosthesis 90.4% (CI 83.9%–97.3%, P = .29). The 5- and 10-year unadjusted survivals were, respectively, 87% (CI 83%–92%) and 68% (CI 62%–76%) for the mechanical valve group and 72% (CI 66%–78%) and 50% (CI 52%–58%) for the bioprosthesis group (P < .01 for both;
Figure 4). Causes of death are listed in
Table 4. Survival of matched patients was compared with survival of all patients aged 50 to 70 years who underwent AVR with either a St Jude mechanical valve (n = 510) or Carpentier-Edwards bioprosthesis (n = 257) between 1990 and 2000 to ensure that the matched patients were representative of the entire cohort. Survivals were similar, indicating a representative sample (
Figure 5).
|
|
|
|
Table 5). In addition, a mechanical valve was protective for late survival (hazard ratio 0.48, CI 0.35–0.67, P < .01).
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionConclusionsReferences |
|---|
Previous studies have shown superior survival for patients with mechanical valves relative to those with bioprosthetic valves. In the Veterans Administration Cooperative Study on Valvular Heart Disease, a prospective, randomized trial, patients who underwent AVR with mechanical valves had greater survival at 15 years than those who received bioprosthetic valves (79% ± 3% vs 66% ± 3%, P = .02).6 In the Edinburgh randomized trial, a survival advantage for mechanical valves was present at 12 years, but at 20 years the survival curves converged, and the small remaining difference was not statistically significant (25% vs 23%, P = .39).7 In a recent retrospective analysis, however, Hanania and associates8 also demonstrated better survival for patients who received mechanical prostheses (46% vs 32%, P = .04).
Other retrospective studies that compared outcomes of patients with current-generation mechanical and bioprosthetic valves have reported no important differences in late survival between the two valve types.9-12 Chan and colleagues10 compared outcomes of patients receiving three different bioprostheses with those of a group of patients who had one of two types of bileaflet mechanical valve.10 They found no difference in valve-related deaths in age categories similar to patients in our study. From what appears to be the same cohort of patients, Prasongsukarn and coworkers12 reported both valve-related and overall mortalities among patients in patients aged 61 to 70 years undergoing AVR; interestingly, in that study, no difference was evident for valve-related late mortality, but there was an overall survival advantage for patients who received a mechanical prosthesis relative to those who received a bioprosthetic aortic valve (4.06% mortality/patient-year vs 5.5% mortality/patient-year, P = .016).
In our study, patients with bileaflet mechanical valves had better survival than did case-matched patients with bioprostheses (hazard ratio 0.46, CI 0.33–0.64). We chose to report overall mortality, rather than cardiac-specific mortality or valve-related mortality, because it was impossible to determine causes of death reliably for all patients. Indeed, without autopsy reports, cause of death (and valve relatedness) is often subjective. In North America, autopsy rates are low, and death certificates are notoriously unreliable for patients with cardiac disease.13-15
Freedom from reoperation on the prosthetic aortic valve was similar between groups, although there was a trend toward more reoperations in the bioprosthesis group. Bioprosthetic valves have a reported freedom from structural valve deterioration ranging between 54% at 15 years to 96% at 12 years,16,17 suggesting that our follow-up may not have been long enough to detect any difference in late reoperation. In addition, our mean patient age was 66 years, which may allow for greater durability of the bioprosthetic valves. In 2004, the life expectancy for Americans at age 65 was 18.7 years (17.1 years for men and 20 years for women).18 In light of the durability of today's bioprosthetic valves, it would appear that many patients who receive a bioprosthetic valve who are younger than 70 years will require a reoperation as an octogenarian, with a potentially increased mortality and morbidity. Whereas aortic valve reoperation mortality as low as 5% has been reported in patients with a mean age of 64 years,19 mortality may be as high as 17.4% in octogenarians.20
An important argument in favor of bioprosthetic valves is the freedom from chronic anticoagulation with warfarin sodium. It should be noted, however, that in our study a significant number of patients with bioprosthetic valves were receiving warfarin sodium or clopidogrel bisulfate at late follow-up. It is important to recognize that patients receive anticoagulation therapy for many reasons, and implantation of a bioprosthetic valve does not ensure that a patient will avoid anticoagulation. These patients are also at risk for atrial fibrillation, transient ischemic attacks, and stroke, which may necessitate long-term anticoagulation. There were more patients in the bioprosthesis group who were receiving clopidogrel bisulfate at last follow-up. This may be because those in the mechanical valve group were already receiving warfarin sodium, which may protect patients from complications related to peripheral vascular disease.
Freedom from major bleeding at 10 postoperative years was better for patients in the bioprosthesis group than for those in the mechanical valve group (93.8% vs 86.5%), although the bleeding rate in the bioprosthesis group was not negligible. It is possible that anticoagulant-related complications may decrease in the future with the introduction of home international normalized ratio monitoring systems that potentially would allow lower levels of anticoagulation in patients with mechanical valves.21 There were no differences in the incidence of stroke between the two groups. Similar to other studies,7,8 there was no difference in late thromboembolism between patients with mechanical valves and those with bioprosthetic valves.
There are several plausible explanations for the findings in this study. Lower survival of patients with bioprostheses may have been due to deaths related to unrecognized bioprosthetic failure. Patients may not have been referred for reoperation or may simply have refused a second or third aortic valve operation because of advanced age or comorbidities. Alternatively, there may have been some hemodynamic benefit for patients with mechanical aortic valves. It is also possible that clinicians implanted mechanical valves preferentially in individuals who generally appeared healthier and thus more likely to outlive a bioprosthetic valve. It is true as well, however, that many active, healthy patients who wished to avoid anticoagulation may have requested and received bioprostheses.
Our study is retrospective, and as such is subject to selection bias. We attempted to minimize patient differences by the case-matching design and by subsequent multivariate modeling techniques. The resulting patient groups were similar with respect to most characteristics and comorbid conditions. A number of surgeons operated during this period, and some had stronger opinions regarding bioprosthetic or mechanical valves for this age group. In a retrospective study, however, surgical bias can never be eliminated or adjusted for retrospectively through propensity matching or other statistical techniques.
|
Conclusions |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionConclusionsReferences |
|---|
|
Footnotes |
|---|
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionConclusionsReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  N. D. Desai, F. McCarthy, W. Moser, W. Y. Szeto, A. Zeeshan, D. Brown, Y. J. Woo, A. Pochettino, P. Moeller, and J. E. Bavaria Durability of Porcine Bioroots in Younger Patients With Aortic Root Pathology: A Propensity-Matched Comparison With Composite Mechanical Roots Ann. Thorac. Surg., December 1, 2011; 92(6): 2054 - 2061. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. Dunning, H. Gao, J. Chambers, N. Moat, G. Murphy, D. Pagano, S. Ray, J. Roxburgh, and B. Bridgewater Aortic valve surgery: Marked increases in volume and significant decreases in mechanical valve use-an analysis of 41,227 patients over 5 years from the Society for Cardiothoracic Surgery in Great Britain and Ireland National database. J. Thorac. Cardiovasc. Surg., October 1, 2011; 142(4): 776 - 782.e3. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 E. A. Ashikhmina, H. V. Schaff, J. A. Dearani, T. M. Sundt III, R. M. Suri, S. J. Park, H. M. Burkhart, Z. Li, and R. C. Daly Aortic Valve Replacement in the Elderly: Determinants of Late Outcome Circulation, August 30, 2011; 124(9): 1070 - 1078. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. Leontyev, M. A. Borger, P. Davierwala, T. Walther, S. Lehmann, J. Kempfert, and F. W. Mohr Redo Aortic Valve Surgery: Early and Late Outcomes Ann. Thorac. Surg., April 1, 2011; 91(4): 1120 - 1126. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. L. d'Arcy, B. D. Prendergast, J. B. Chambers, S. G. Ray, and B. Bridgewater Valvular heart disease: the next cardiac epidemic Heart, January 15, 2011; 97(2): 91 - 93. [Full Text] [PDF]
|
|
|
|
|
|
M. E. Halkos, P. Kilgo, O. M. Lattouf, J. D. Puskas, W. A. Cooper, R. A. Guyton, and V. H. Thourani The Effect of Diabetes Mellitus on In-Hospital and Long-Term Outcomes After Heart Valve Operations Ann. Thorac. Surg., July 1, 2010; 90(1): 124 - 130. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. T. Nguyen, F. Delahaye, J.-F. Obadia, X. Duval, C. Selton-Suty, J.-P. Carteaux, B. Hoen, F. Alla, and for the AEPEI study group Aortic valve replacement for active infective endocarditis: 5-year survival comparison of bioprostheses, homografts and mechanical prostheses Eur J Cardiothorac Surg, May 1, 2010; 37(5): 1025 - 1032. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 P. Stassano, L. Di Tommaso, M. Monaco, F. Iorio, P. Pepino, N. Spampinato, and C. Vosa Aortic Valve Replacement: A Prospective Randomized Evaluation of Mechanical Versus Biological Valves in Patients Ages 55 to 70 Years J. Am. Coll. Cardiol., November 10, 2009; 54(20): 1862 - 1868. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
P. H. Stone Current Selection of Optimal Prosthetic Aortic Valve Replacement in Middle-Aged Patients: Still Dealer's Choice J. Am. Coll. Cardiol., November 10, 2009; 54(20): 1869 - 1871. [Full Text] [PDF]
|
|
|
|
|
|
S. H. Rahimtoola The Year in Valvular Heart Disease J. Am. Coll. Cardiol., May 19, 2009; 53(20): 1894 - 1908. [Full Text] [PDF]
|
|
|
|
 |
|
 A. Vahanian, B. Iung, L. Pierard, R. Dion, and J. Pepper CHAPTER 21 Valvular Heart Disease ESC Textbook of Cardiovascular Medicine, January 1, 2009; 2(1): med-9780199566990-chapter - med-9780199566990-chapter. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Brown and H. Schaff Reply to the Editor: J. Thorac. Cardiovasc. Surg., October 1, 2008; 136(4): 1102 - 1102. [Full Text] [PDF]
|
|
|
|
|
|
A. Cannata, C. F. Russo, and C. Taglieri Aortic valve replacement dilemma: mechanical or biological prosthesis? J. Thorac. Cardiovasc. Surg., October 1, 2008; 136(4): 1101 - 1102. [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 |
