| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Ann Thorac Surg 2007;83:549-557
© 2007 The Society of Thoracic Surgeons
Division of Cardiovascular Surgery, Department of Surgery, University of British Columbia, St. Paul’s Hospital, Vancouver General Hospital, and Royal Columbian Hospital, Vancouver, Canada
Accepted for publication July 18, 2006.
* Address correspondence to Dr Jamieson, 486 Burrard Building, St. Paul’s Hospital, 1081 Burrard Street, Vancouver, BC V6Z 1Y6, Canada (Email: wrej{at}interchange.ubc.ca).
Presented at the Forty-second Annual Meeting of The Society of Thoracic Surgeons, Chicago, IL, Jan 30–Feb 1, 2006.
 |
Abstract |
|---|
 Top Abstract IntroductionPatients and MethodsResultsCommentDiscussionAcknowledgmentsReferences |
|---|
METHODS: From 1994 to 2000, 1,372 patients implanted with three currently marketed aortic bioprostheses, Medtronic Mosaic (Medtronic, Inc, Minneapolis, MN) (415 patients), Carpentier-Edwards SAV (462), and Carpentier-Edwards PERIMOUNT (495) (Edwards Lifesciences, Irvine, CA), with a mean age of greater than 70 years were evaluated. Patient populations were comparable, inclusive of concomitant coronary artery bypass grafting (CABG) for the overall populations and for patients greater than 70 years.
RESULTS: There were 37 thromboembolic (TE) events: major TE, 14; reversible ischemic neurologic deficit (RIND), 12; and minor TE, 11. There were 4 TE deaths. Multivariate (stepwise logistic regression) analysis revealed no predictive risk factors for overall TE. For the combination of major TE plus RIND there were two predictive risk factors with analysis of 12 risk variables: preoperative cerebrovascular accident (odds ratio [OR] 4.45, 95% confidence interval [CI] 1.17 to 16.87, p = 0.028); and concomitant CABG (OR 3.19, 95% CI 1.16 to 8.76, p = 0.025). Neither anticoagulant nor antiplatelet therapies gave significant protection.
CONCLUSIONS: There does not appear to be an indication for routine antithrombotic management. The study supports the potential use of antithrombotic therapy for comorbidities of preoperative cerebrovascular accident and concomitant CABG but not atrial fibrillation, left ventricular dysfunction, or elderly age greater than 70 years. Vascular burden and advanced age are likely contributing factors to these independent predictors. There may still be a need for, or at least consideration of, a randomized trial for AVR with bioprostheses.
|
Introduction |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsCommentDiscussionAcknowledgmentsReferences |
|---|
The European Society of Cardiology (ESC) has produced two consensus documents outlining their recommendations regarding anticoagulation. In 1995, they recommended an anticoagulant regimen of international normalized ratio (INR) 2.5 to 3.0 for three months for bioprosthetic aortic valve replacement (AVR) in sinus rhythm [1]. In the 2005 publication of the ESC, the recommendation remained unchanged but recognized the widespread use of aspirin as an alternative to anticoagulation for the first three months [8]. In 1998, the British Society of Haematology recommended no anticoagulation for aortic bioprostheses in sinus rhythm [2, 3]. The American College of Cardiology/American Heart Association (ACC/AHA) consensus report of 1998 recommended early heparin and anticoagulation with an INR of 2.5 to 3.5 for the first three months after bioprosthetic AVR [4]. The ACC/AHA Practice Guidelines published in 2006 [5] are now recommending aspirin alone in patients with AVR bioprostheses and no risk factors, while aspirin combined with anticoagulation in high-risk patients with AVR bioprostheses. The American College of Chest Physicians (ACCP) in 2001 [6] recommended an INR target of 2.0 to 3.0 for bioprosthetic AVR in sinus rhythm. In 2004, the Canadian Cardiovascular Society consensus document on "Surgical management of valvular heart disease" recognized both aspirin and coumadin as anticoagulation regimes but with an evidence class of CII for both modes of management [7].
The Cardiothoracic Surgery Network (CTSNet) reported [10, 11], in a survey conducted of 726 surgeons in 2004, that 79% were aware of the 1998 ACC/AHA guidelines with 21% being unaware. Furthermore, 58% of the respondent surgeons felt that antiplatelet therapy should be the standard of care after AVR with bioprostheses in patients without comorbidities and 68% felt that coumadin-warfarin was no longer the standard of care. The 2006 ACC/AHA guidelines are now supportive of this surgical practice opinion.
There are several single center publications in the literature [12–16], five recommending against anticoagulation and one recommending anticoagulation. The current study was conducted to determine if antithrombotic therapy was protective against TE with aortic bioprostheses 30 days or less after AVR.
|
Patients and Methods |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsCommentDiscussionAcknowledgmentsReferences |
|---|
From 1994 to 2000 inclusive, at the affiliated teaching hospitals (St. Paul’s Hospital, Vancouver General Hospital, and Royal Columbian Hospital) of the University of British Columbia, 1,372 patients were implanted with currently marketed bioprostheses. The Medtronic Mosaic porcine (MM) (Medtronic Inc, Minneapolis, MN) was implanted in 415 patients (mean age 70.4 ± 10.7 years; range, 26 to 89 years), the Carpentier-Edwards SAV porcine (CE-SAV) (Edwards Lifesciences, Irvine, CA) in 462 patients (mean age, 72.7 ± 8.3 years; range, 32 to 89 years), and the Carpentier-Edwards PERIMOUNT pericardial (CE-P) (Edwards Lifesciences) in 495 patients (mean age, 74.4 ± 8.4 years; range, 25 to 94 years).
The characteristics of the total population and the individual valvular populations are detailed in Table 1. The descriptive characteristics of the total population are summarized in Table 2 by the presence or absence of various potential risk factors of TE. The twelve risk factors considered were the following: atrial fibrillation, left ventricular dysfunction (low ejection fraction < 0.3), previous cerebrovascular accident (CVA), carotid disease-endarterectomy, peripheral vascular disease, ventricular arrhythmias, age greater than 70 years, diabetes mellitus, gender, concomitant coronary artery bypass grafting (CABG), previous valve surgery, and previous CABG. The antithrombotic therapy during the study era was anticoagulation with a target INR of 2.5 to 3.5 and (or) acetylsalicylic acid (ASA) at a dosage of 81 mg to 325 mg.
|
|
The early TE events (
30 days) are presented as crude event rates and number of events. The mean age was presented ± standard deviations. The statistical tests used to assess continuous and categoric variables are designated on the appropriate tables.
The multivariate stepwise logistic regression (without time in the model) was used to determine the independent predictors of TE overall and a subset of TE (major TE and RIND) using the 12 potential risk factors previously documented above in this section. Multivariate stepwise logistic regression was also used to assess limited risk factors from the univariate model and those of clinical importance. The multivariate modeling was also used to assess predictive risk factors in the total population without the patient subset with atrial fibrillation. The independent risk factors are presented as odds ratios with 95% confidence limits and statistical significance level.
|
Results |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsCommentDiscussionAcknowledgmentsReferences |
|---|
30 days) TE events by bioprosthesis type are illustrated in Table 3. There were four TE deaths within the 30 or less day period: MM (2), CE-SAV (1), and CE-P (1). There were six hemorrhagic events related to antithrombotic therapy.
|
The results of the multivariate stepwise logistic regression analysis for overall TE (37 events; 2.7%) and for the combination of major and RIND events (26; 1.8%) are detailed in Table 4. There were no predictors of TE for the overall 37 events. There were, however, two predictors for major and RIND events; preoperative CVA (odds ratio [OR] 4.45, p = 0.028) and concomitant CABG (OR 3.19, p = 0.025). Age greater than 70 years was a predictor by univariate analysis (OR 3.95, p = 0.026) but not by multivariate analysis (OR 3.29, p = 0.060). Of the 26 major and RIND events, 3 occurred with preoperative CVA and 21 with concomitant CABG. Only 2 accompanied atrial fibrillation, 1 with ejection fraction less than 0.35, while 23 were greater than 70 years of age. Only 3 occurred with previous CABG.
|
|
2 statistics had a p value less than 0.25. These were atrial fibrillation, preoperative CVAs, age greater than 70 years, concomitant CABG, and previous CABG. There were no independent predictors of overall TE (n = 37), while for TE major plus RIND (n = 26) only concomitant CABG (OR 3.2, p = 0.022) was predictive. To assess the role of gender and medications the modeling was repeated with atrial fibrillation, preoperative CVA, age greater than 70 years, gender, and discharge medications (anticoagulant [AC] or AC + antiplatelet [AP], AP, or none). There were no predictors of overall TE (n = 37), while for TE major plus RIND (n = 26) – preoperative CVA (OR 3.62, p = 0.044) and age greater than 70 years (OR 3.97, p = 0.026).
To the above modeling, concomitant CABG was added to assess six independent predictors (Table 5). The only independent predictor was concomitant CABG (OR 3.10, p = 0.027) replacing preoperative CVA (OR 3.25, p = 0.068) and age greater than 70 years (OR 3.04, p = 0.077).
|
The study was extended to determine the number of TE events that occurred between 30 or less days and 90 or less days. There were only five events during this period; one major TE event, one RIND and three minor TE events. The modeling for independent predictors for overall TE (n = 42) revealed previous CABG as marginally predictive (OR 3.13, p = 0.051) and valve type CE-P (OR 0.392, p = 0.029). When univariate descriptive statistics with p less than 0.025 were analyzed, CE-P was predictive and protective (OR 0.445, p = 0.048). When the modeling was conducted only for major plus RIND events (n = 28), the only predictive variable was preoperative CVA (OR 4.49, p = 0.026).
|
Comment |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsCommentDiscussionAcknowledgmentsReferences |
|---|
The ESC has reported twice over the decade; first in 1995 [1] and again in 2005 [8]. The ESC recommended anticoagulation at an INR target of 2.5 to 3.0 for three months after surgery. The 1998 ACC/AHA guidelines recommend [4] anticoagulation for three months after bioprosthetic AVR with an INR target of 2.5 to 3.5. The 2006 ACC/AHA practice guidelines [5] are now recommending aspirin alone in patients with AVR bioprostheses and no risk factors, while aspirin combined with anticoagulation in high-risk patients with AVR bioprostheses. The American College of Chest Physicians in 2001 [6] also recommended anticoagulation for three months but with a reduced INR target of 2.0 to 3.0. The only consensus document to completely recommend no anticoagulation was the British Society of Haematology, reporting in 1998 [2, 3]. The Canadian Cardiovascular Society Consensus on Surgical Management of Valvular Heart Disease acknowledged that there was extensive use of ASA in the surgical community and that there was evidence lacking to recommend anticoagulation or ASA therapy for bioprosthetic AVR for patients with sinus rhythm and no other risk factors [7].
There is a general consensus that anticoagulation should be utilized in the early postoperative period and in the long term, for bioprostheses with accompanying risk factors of TE. The ACC/AHA acknowledged risk factors of atrial fibrillation, left ventricular dysfunction, previous TE, and hypercoagulable condition [4, 5]. The British Society of Haematology stated the risk factors of TE in a slightly different categorization including previous TE, left atrial thrombus, marked cardiomegaly, heart failure, dilated left atrium, or spontaneous echocontrast [2, 3].
Acetylsalicylic acid as an alternative to anticoagulation in patients with sinus rhythm, without demonstrable risk factors, was demonstrated in the survey conducted by the CTSNet and reported in 2004 [10, 11]. The survey was conducted of 728 surgeons and the general recommendation was that antiplatelet therapy should be the standard of care in AVR patients without comorbidities for bioprostheses and that coumadin-warfarin should no longer be considered the standard of care. The 2006 ACC/AHA Practice Guidelines [5] are now recommending aspirin alone in patients with AVR bioprostheses and no risk factors, while aspirin combined with anticoagulation in high-risk patients with AVR bioprostheses.
The 1998 ACC/AHA guidelines [4] were formulated from weak data without a clear demonstration of risk to benefit ratio. The current (2006) ACC/AHA guidelines [5] have revised the recommendations to aspirin alone for bioprosthetic AVR and aspirin + anticoagulation for bioprosthetic AVR in the presence of high-risk factors inclusive of atrial fibrillation, previous TE, left ventricular dysfunction, and hypercoagulable conditions. These current (2006) guidelines are classified as "evidence level C"; that is, only consensus opinion of experts, case studies, or standard-of-care. The study that influenced the original (1998) ACC/AHA guidelines was one of two publications from the Mayo Clinic. Heras and colleagues [12] reported in 1995 that the TE risk was especially high for 90 days after AVR with bioprostheses and that the rate increased with lack of anticoagulation. These authors demonstrated that the greatest risk period was between one and 10 days after the surgery with reduced risk between 10 and 90 days. They concluded from their study that anticoagulation reduces TE and appears to be indicated in all patients as early as possible for a period of three months after bioprosthetic AVR.
There have been recommendations for reduced INR target levels to 2.0 to 3.0 from the ACCP and from the Canadian Cardiovascular Society Consensus document on surgical management of valvular heart disease [6, 7]. The first report that acknowledged the effectiveness of reduced INR target levels with reduction of hemorrhagic complications was from Turpie and colleagues [18] in 1988. They evaluated a randomized trial of bioprostheses in the aortic position with randomization of INR to 2.0 to 2.5 and 2.5 to 4.0. They demonstrated equivalence in prophylaxis in TE but a significantly reduced incidence in hemorrhagic events with the lower INR target. There is now general consensus that the INR target for aortic valve bioprostheses can be lowered safely to a level of INR 2.0 to 3.0.
The majority of publications since the 1998 ACC/AHA guidance document have recommended that anticoagulation with warfarin-coumadin is not necessary in patients with sinus rhythm and absence of risk factors [13–16]. Gherli and colleagues reported in 2004 [13], in a prospective study that there were no advantages to anticoagulation therapy compared with antiplatelet therapy with regard to early cerebral ischemic events, bleeding, and survival. These authors concluded that there was currently no evidence to support the fact that warfarin- coumadin is more effective than ASA [13]. Babin-Ebell and colleagues [14] also concluded that there was no benefit to anticoagulation over antiplatelet therapy. Moinuddeen and colleagues [15] reported in 1998 that, from a retrospective study, early anticoagulation after AVR with a bioprosthesis confers no advantage in the prevention of early cerebral ischemic events. These authors concluded that early anticoagulation after bioprosthetic AVR was unnecessary [15].
The second report from the Mayo Clinic was in 2005 and in this report Sundt and colleagues [16] evaluated the 90-day incidence of cerebrovascular events with and without anticoagulation. They concluded that postoperative CVA occurred in 2.4% of the patients who received anticoagulation and 1.9% of patients who did not receive anticoagulation [16]. These authors stated that postoperative CVA was unrelated to anticoagulation. These authors also demonstrated that there were no variables predictive of bleeding, whether patients were on anticoagulation or not. This study concluded that anticoagulation did not appear to protect against neurologic events, but the limitation of the study was that the role of antiplatelet therapy was not addressed.
The current study incorporated AVRs performed with bioprostheses at the University of British Columbia affiliated teaching hospitals between 1994 and 2000. The stimulus for the study was related to the results of the experience of the center with the Medtronic Mosaic porcine bioprosthesis that was presented to the Society for Heart Valve Disease in 2003 and subsequently published in the Journal of Heart Valve Disease [19]. There was a concern that the TE rates were elevated with the MM porcine bioprosthesis but the study of the results were comparable with the CE- SAV porcine and the CE-P pericardial bioprostheses. In 1989, the University of British Columbia cardiac surgeons embarked upon antiplatelet therapy as routine management for antithrombotic therapy for patients at low risk of TE. Patients continued to be anticoagulated for risk factors of TE including chronic atrial fibrillation. This practice pattern continued through the timing of this study.
The findings of the study give further information for use of anticoagulation with aortic bioprostheses. It has become apparent from this study that predominantly concomitant CABG and preoperative CVA were the independent predictors of postoperative TE, primarily for major events and RIND events. Peripheral vascular disease was not considered an etiological factor for events occurring with concomitant CABG and preoperative CVA because only one TE event occurred in patients with documented peripheral vascular disease. The interesting finding of this study was that considered risk factors were not predictive of overall TE. Atrial fibrillation and left ventricular dysfunction as defined by an ejection fraction less than 0.35 were not predictive of TE. There was a definite trend for patients greater than 70 years to be predictive of TE. The presence of anticoagulation with or without antiplatelet therapy at discharge was 11% in the total population while in atrial fibrillation it was 37%; ejection fraction less than 0.35 was 15.5% and patients greater than 70 years was 12.5%. The preoperative CVA group was managed with anticoagulation with or without antiplatelet therapy at discharge in 17% and with only antiplatelet therapy in 64% of the group. There were 26 major and RIND events and three of these occurred in patients with a history of preoperative CVA and 21 accompanied concomitant CABG. Twenty-three of the 26 patients were greater than 70 years of age. Patients in this study had a mean age of 72.6 years of age.
The predictive modeling for independent predictors of major TE (major + RIND) in the various assessments revealed that the predictors were preoperative CVA and age greater than 70 when concomitant CABG was not in the model; when concomitant CABG was in the model, concomitant CABG was the only independent predictor. These findings are more suggestive of vascular disease burden than of actual predictors of major TE. This may be the case for concomitant CABG was performed in 59% of the patients greater than 70 years of age who would be considered to have greater vascular disease burden. In all of the models discharge medications, both anticoagulation and antiplatelet therapy, showed nonsignificant evidence of being protective. The unknown in this study is whether or not antithrombotic therapy would have a role in controlling the apparent vascular disease burden that may be the sole etiologic factor of postoperative TE after AVR with bioprostheses. An interesting component of the findings was that anticoagulation was utilized in only 37% of the patients with atrial fibrillation and that when the modeling was conducted with all atrial fibrillation patients removed from the analysis, the predictors continued to be preoperative CVA and age greater than 70 years. The elimination of atrial fibrillation was not evaluated with the assessment of concomitant CABG as a variable, which displaced preoperative CVA and age greater than 70 years in the other documented models.
The characteristics of the three patient populations with the three currently marketed bioprostheses deserve further comment. The patients with the three bioprostheses were undifferentiated with regard to previous CABG, concomitant CABG, and previous valvular surgery. The status of left ventricular function was undifferentiated between the bioprostheses but there was a definite difference in numbers between the levels of ejection fraction. There were predominantly two differentiating factors among the patient populations and these are related to gender and preoperative CVA. There was a preponderance of preoperative CVA in the patients who received the CE-P bioprosthesis and this patient population had a higher incidence of female patients. The reason for the higher incidence of female patients is likely related to the presumed superior hemodynamics with the pericardial bioprosthesis. In 2000, our center reported that there was no difference in the hemodynamic performance between these three bioprostheses and consequently the practice of utilizing the pericardial valve in the small aortic root was reconsidered [20]. In all the modeling of major TE 30 days or less, valve type was not predictive statistically; however, when the modeling for major TE was within 90 days the valve type (namely CE-P) became predictive in providing protection. This finding is not easily explained, for the percentage of patients experiencing major TE plus RIND was not clinically different (CE-SAV 1.7% [8 of 462], MM 2.2% [9 of 415], and CE-P 1.8% [9 of 495]. The only clinically relevant finding was the percentage of overall TE, inclusive of major RIND and minor, where the CE-P rate was 2.0% while the CE-SAV and MM rates were 3.0% and 3.1%, respectively.
This study specifically addresses the role of presumed TE risk factors. From this evaluation, we have determined that atrial fibrillation and compromised left ventricular function were not predictive of TE in our series. The two predominant predictors were concomitant CABG and preoperative CVA. Patients greater than 70 years of age showed a strong indication toward prediction of TE. Previously reported studies particularly addressed overall TE in analysis and risk assessment. In this study, the assessment of independent predictors of overall TE was negative. It was only when major TE events inclusive of RIND were assessed in the risk assessment did we identify predictors of major events. It is conceivable that minor TE events are related to thrombi on the sewing ring prior to complete endothelialization and have a different etiology than major events.
In conclusion, the recommendations of this study are threefold. First, there appears to be no definite indication for routine anticoagulant management with AVR with bioprostheses. Second, anticoagulation may be necessary for the comorbidities of preoperative CVA and concomitant CABG representing vascular disease burden. Third, several of the considered risk factors of TE did not play a role in our series; namely, atrial fibrillation and left ventricular dysfunction. It seems apparent that the cardiac surgical community is probably correct in their resistance to administer routine antithrombotic therapy with anticoagulation. Consideration should be given of these results in the potential for the performance of a randomized trial to address completely the issue of antithrombotic therapy for aortic valve bioprostheses.
|
Discussion |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsCommentDiscussionAcknowledgmentsReferences |
|---|
In the recent past, Donald Rumsfeld made some comments using "management talk" and MBA type phrases in saying: "As we know, there are known knowns. There are things that we know we know. We also know there are known unknowns. That is to say, we know there are some things we do not know." He went on to say: "But there are also unknown unknowns, the ones we don’t know we don’t know" (DoD briefing, 2/12/02). His last statement seemed to get a lot of people upset. Thus, I am wondering if the dilemma presented today is a similar problem?
We have published guidelines on the role of anticoagulation post bioprosthetic aortic valve replacement from multiple sources, the most recent, being the ACC/AHA, which were published in 1998. One would think that this issue has been resolved and that following these guidelines would be the way we should go with this issue. However, your study presented today as well as previous studies give a different conclusion. Moinuddeen and colleagues with a retrospective study from Yale published in Circulation in 1998; Gary and associates from the University of Parma did a prospective study published in Circulation in 2004; and Sundt and colleagues from the Mayo Clinic published in the Journal of Thoracic and Cardiovascular Surgery in 2005; all reached the same conclusions: that anticoagulation therapy was unwarranted after bioprosthetic aortic valve replacements.
This prompted me to look at our own experience from The Lankenau Hospital over the past 3
years. There were 443 aortic valve replacements, of which 91% were bioprosthetic valves; and 72% of those were isolated aortic or aortic valve and CABG procedures. Our cerebrovascular accident rate was 0.7% and our TIA/RIND was 3.1%. There were no major thromboembolic events, and only 5% of the patients received Coumadin. It is not our practice to give anticoagulation therapy to patients receiving a bioprosthetic aortic valve in the postoperative period at The Lankenau Hospital.
So my three questions to you are the following. First, the guidelines, at least from the ACC/AHA, seemed to have been profoundly influenced by a previous Mayo Clinic study that looked at patients from 1975 to 1982. Your study was completed approximately 12 years later when patient data were accrued and analysis occurred. Based on your present experience and knowledge in this field, what if any role would current generation of valve designs as well as cardiopulmonary bypass technology play in decreasing or in influencing a low rate of thromboembolic events in your patients?
Second, your present study cites the CTSNet survey of 726 surgeons; most of the respondents were from the USA or Europe. This CTSNet survey also indicated that approximately 39% of those surgeons do less than 20 aortic valves per year and approximately two-thirds do less than 30 valves per year. I believe these findings are also consistent with previous STS Workforce surveys. So given that most surgeons may not have as extensive an experience as you do; and some like myself, whose practice is in the state of Pennsylvania in an environment of public reporting, I would like to know in a straightforward way what are your own guidelines and how do you currently treat your patients after aortic valve replacement?
Lastly, asking a physician to participate in a prospective randomized trial—when clearly, by the survey and studies just mentioned, many if not most surgeons do not believe and/or follow the guidelines—may not be feasible. A recent article and thoughtful editorial in the New England Journal of Medicine looked at the issue of potentially changing current clinical recommendations or ingrained medical practice (N Engl J Med 2006;354:353–65 and 329–31). Might not an alternative approach be to use good quality data such as from surgeons who participate in the STS National Database, who are now listed in the back of our Program Book by a heart symbol (
) next to their names, and pooling these data from multiple sources and then using modern statistical tools such as propensity score matching to evaluate this question rather than a prospective randomized trial?
I would like to thank again the Society and Dr Jamieson for the privilege of discussing this paper.
DR JAMIESON: I appreciate your evaluation and thoughtful comments. The type of prosthesis we utilized potentially could be a factor. We were an investigating center for the Medtronic Mosaic starting in 1994, and this study was initiated after we looked at our experience with the Mosaic, and that is published in the Journal of Heart Valve Disease. We were concerned that the rates might be higher with the Mosaic. We then evaluated our other two prostheses and found no difference.
The SAV valve, which has only been available in the United States since 2002, was actually available to us since 1982, and we have recently published our 18 year experience. The pericardial valve, though, was available in the United States since well in the 1980s. So whether that is a factor or not I am uncertain.
Our general experience has been since 1989, and this included other centers in Canada, where we utilized primarily antiplatelet therapy for those patients in sinus rhythm. Our general current guidelines remain the same except for this consideration, that we have currently found the specific indications for surgery.
I certainly agree that a more advanced study utilizing potentially the STS database, and since we do have data within 30 days in the STS database and certainly could do propensity matching on patients, would be a very, very valid study and I would certainly support that. We have given consideration, ourselves and other centers in Canada showing this data, whether or not a randomized trial might be considered, and at the present time we have seen no specific indications and other centers in Canada have also ruled out the potential of a randomized trial and the benefit of it.
DR WILLIAM A. BAUMGARTNER (Baltimore, MD): Eric, terrific presentation, and I think many of us appreciate you bringing this to our attention. I have just two questions. Are we at a point now where the ACC/AHA guidelines ought to be modified based upon most recent data, even though it is not randomized? We have already heard in other presentations, and they all come to the same conclusion. I wonder if it is now time that our STS representatives consider bringing this to the table again?
The reason not to anticoagulate patients is because of bleeding complications, and I wondered if, in the group of patients that did have anticoagulation, what is your incidence of bleeding complications and pericardial tamponade?
DR JAMIESON: In answer to your first question, I believe the ACC/AHA guidelines are under review, and I am not sure when they will be published. We published Canadian guidelines in 2004, and we really were unable to differentiate and make a definite recommendation; both aspirin and ASA [acetylsalicylic acid] were a C2 classification. We did not have a serious problem with hemorrhage related to the surgical procedure in this case, and that rate was well below 1%. In 1989, when we embarked upon this utilization of only antiplatelet therapy, that was particularly the reason why we did it. We felt we had more morbidity from tamponade and postop surgical bleeding after the patient went home from the hospital than we did from thromboembolic events.
|
Acknowledgments |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsCommentDiscussionAcknowledgmentsReferences |
|---|
|
References |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsCommentDiscussionAcknowledgmentsReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  P. Pibarot and J. G. Dumesnil Prosthetic Heart Valves: Selection of the Optimal Prosthesis and Long-Term Management Circulation, February 24, 2009; 119(7): 1034 - 1048. [Full Text] [PDF]
|
|
|
|
 |
|
 A. Colli, J.-P. Verhoye, A. Leguerrier, and T. Gherli Anticoagulation or antiplatelet therapy of bioprosthetic heart valves recipients: an unresolved issue Eur. J. Cardiothorac. Surg., April 1, 2007; 31(4): 573 - 577. [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 |
