Self-Management of Oral Anticoagulation Therapy Improves Long-Term Survival in Patients With Mechanical Heart Valve Replacement

Transcription

1 CARDIOVASCULAR ORIGINAL ARTICLES: CARDIOVASCULAR Self-Management of Oral Anticoagulation Therapy Improves Long-Term Survival in Patients With Mechanical Heart Valve Replacement Heinrich Koertke, MD, Armin Zittermann, PhD, Otto Wagner, and Reiner Koerfer, MD, PhD Department of Cardiothoracic Surgery, Heart Center North-Rhine-Westphalia Bad Oeynhausen, Clinic of the Ruhr University Bochum, Germany Background. The Early Self-Controlled Anticoagulation Trial has demonstrated that in patients with mechanical heart valve replacement self-management of oral anticoagulation results in less major thromboembolic events than conventional measurement by the general practitioner. However, the effects of selfmanagement on long-term survival are currently not known. Methods. Nine hundred thirty patients participated in a follow-up study of the aforementioned trial (488 from the self-management group and 442 from the conventional group). Long-term survival was assessed 12 years after the study began using the intent-to treat analysis as well as the per protocol analysis. Univariate and multivariate analyses were performed in order to assess independent predictors of survival. Results. In total, the 930 patients accrued 8,315 patientyears of observation. During follow-up, 236 patients died. According to the intent-to treat analysis, 10-year survival was 76.1% in the conventional group and 84.5% in the self-management group. The corresponding values for the per protocol analysis were 67.7% and 80.6%, respectively. Age, kind of valve surgery, and study group were independent predictors of survival. Self-management of oral anticoagulation increased long-term survival by 23% (intent-to-treat analysis) and 33% (per protocol analysis), respectively, compared with conventional measurement by the general practitioner. Possible reasons for these advantageous results in the self-management group are fewer thromboembolic events due to a higher percentage of international normalized ratio values lying in the target range compared with the conventional group. Conclusions. Data indicate that self-management of oral anticoagulation is a promising strategy in order to increase long-term survival in patients with mechanical prosthetic valves. (Ann Thorac Surg 2007;83:24 9) 2007 by The Society of Thoracic Surgeons Patients with mechanical prosthetic heart valves require long-term anticoagulation therapy for prevention of systemic thromboembolism. Based on positive experience with the self-management of diabetes mellitus and hypertension [1, 2], it has been assumed that a significant percentage of patients on anticoagulation therapy will be able to perform testing and adjustment of their international normalized ratio (INR) values self-responsible [3]. Meanwhile, some studies have demonstrated that the patients of self-management groups had 6% to 18% more INR values lying within the target range compared with conventionally treated patients [4 8]. However, in most of these earlier studies patient selection was restrictive, and the total number of included patients was small, ranging from 46 [4] to 165 [5]. Consequently, no data on Accepted for publication Aug 11, Address correspondence to Dr Koertke, Herz- und Diabeteszentrum NRW, Georgstr.11, Bad Oeynhausen, Germany; hkoertke@ hdz-nrw.de. complication rates and cumulative survival were available from these studies. The Bad Oeynhausen concept of INR self-management is based on the assumption that generally each patient who depends on anticoagulation therapy for a longer period of time or even lifelong is able to learn INR self-management [3]. The Early Self-Controlled Anticoagulation Trial I (ESCAT I) study is a large, randomized, prospective study with 1,155 patients. The investigation was already initiated in February 1994 [9]. This study has demonstrated that 79% of the INR values were within the target range of 2.5 to 4.5 in the self-management group compared with only 65% in the group who had their INR values determined by a general practitioner [10]. Moreover, the ESCAT I study showed that the INR selfmanagement resulted in less major thromboembolic events compared with the conventionally treated group [10]. Here, we present data demonstrating that INR self-management is also able to improve long-term survival of patients with mechanical heart valve prostheses compared with the conventionally treated group by The Society of Thoracic Surgeons /07/$32.00 Published by Elsevier Inc doi: /j.athoracsur

2 Ann Thorac Surg KOERTKE ET AL 2007;83:24 9 INR SELF-MANAGEMENT AND LONG-TERM SURVIVAL Patients and Methods Patients A total number of 930 patients (65.5% males) participated in the ESCAT I follow-up study. At follow-up, age ranged from 28 to 87 years (mean age, 70 years). Patients had undergone a mechanical aortic, mitral, or double heart valve replacement. Three hundred and forty-four Carbomedics valves (Carbomedics, Austin, TX), 381 St Jude Medical valves (St. Jude Medical GmbH, Nürnberg, Germany), and 205 Medtronic Hall valves (Medtronic GmbH, Düsseldorf, Germany) had been implanted. All patients had given written informed consent to the study procedures. The study protocol had been approved by the Ethics Committee of the Ruhr University Bochum, Germany. Study Procedures and Parameters The design of the ESCAT I study is described in detail elsewhere [9, 10]. Briefly, patients were randomly assigned to the group that performed INR selfmanagement (designated SEM group) or the group that had their INR values measured by a general practitioner (designated controlled [CON] group). On the sixth or seventh postoperative day, the patients of the SEM group were instructed in INR self-management. In contrast, patients who had their INR values measured by the general practitioner were not instructed in the principles of INR measurement in order to mirror the traditional way of INR management in Germany. In the CON group, the anticoagulant dosage was advised and controlled by a general practitioner. Once a month, all participants of the SEM group sent the collected coagulation data to the study center for further analysis. The INR values of the CON group were sent to the study center whenever INR testing was performed by the general practitioner. All patients of the ESCAT study received the anticoagulant phenprocomoun. Every 6 months, an outpatient control was carried out at the study center. All together, 41,792 INR values of the SEM group and 8,905 INR values of the CON group were sent to the study center for further analysis. Two years after valve implantation, the patients stopped sending their INR values to the study center. Thereafter, patients of the SEM group were encouraged to continue with INR self-management. Because, until a short time ago, a therapeutic INR range from 2.5 to 4.5 had been regarded as necessary to prevent thromboembolism, this was the target range for both study groups 25 CARDIOVASCULAR Table 1. Baseline Characteristics of the Study Groups Characteristics Conventional Group N 442 Self-Management Group N 488 Significance Body mass index (kg/m 2 ) Ejection fraction NYHA functional class (%): I II III IV Hypertension (%) Atrial fibrillation (%) Smokers (%) Primary diagnosis: Aortic stenosis Coronary heart disease Aortic insufficiency Mitral insufficiency Others Unknown Valve replacement: Aortic valve Mitral valve Double valve Valve provider: Carbomedics St. Jude Medical Metronic Hall Concomitant procedures (%): CABG Others CABG coronary artery bypass grafting.

3 CARDIOVASCULAR 26 KOERTKE ET AL Ann Thorac Surg INR SELF-MANAGEMENT AND LONG-TERM SURVIVAL 2007;83:24 9 during the entire follow-up period. Enrollment of patients started in February 1994 and was finished in September In April and May 2006, we called the patients by telephone in order to assess whether or not they were still performing INR self-management or INR measurements by the general practitioner. Moreover, we assessed survival of the patients. In those patients who died during follow-up, we requested the medical report from the general practitioner, the emergency hospital, or the life insurance in order to determine the cause of death. Causes of death were classified into intraoperative deaths, evidently anticoagulation-related (such as thromboembolism or bleeding), cardiac-related (such as heart failure, myocardial infarction, cardiogenic shock, atrial fibrillation, or cardiac arrest), others (such as infection, inflammation, asthma bronchiale, malignancy, or suicide), and unknown. Statistics Statistical evaluations were performed with the Statistical Package for Social Sciences version 11 (SPSS, Chicago, IL). Categoric variables were reported using the number (N) and percent of observations. Continuous variables were expressed as mean values with standard error. For comparative evaluations, the Fisher exact test, the 2 test, the Friedman test, and the Mann-Whitney U test were used. Survival rates were calculated with the Kaplan-Meier product-limit estimator. The log-rank test was used to test for differences in survival rates of specific subgroups. Moreover, we used the univariate Cox regression model in order to assess associations of factors to long-term survival. Every univariate variable showing marginal significance (p 0.15) was then tested in a Cox multivariate model and removed stepwise if no significant influence was proved. Variables tested were as follows: age, sex, body height, body weight, left ventricular ejection fraction, New York Heart Association functional class, hypertension, atrial fibrillation, valve position, study group, and smoker or nonsmoker. For comparative Fig 2. Kaplan Meier survival estimates for patients with self-management (SEM group) of their international normalized ratio (INR) values in comparison with patients who had their INR values measured by the general practitioner (controlled [CON] group). analysis of long-term survival, the intent-to-treat method was used. Moreover, a per protocol analysis was performed. A p value less than 0.05 was considered statistically significant. Results Baseline characteristics of the two study groups are listed in Table 1. The percentage of patients with mitral valve Table 2. Causes of Deaths of the Study Groups Cause of Death Conventional Group (N 442) Self-Management Group (N 488) Fig 1. Distribution of the international normalized ratio (INR) values of patients with INR self-management (SEM group) and of patients who had their INR values measured by the general practitioner (controlled [CON] group). Intraoperative death 9 3 Potentially anticoagulation 7 8 related: Cerebral or intestinal 2 5 bleeding Embolism 0 1 Stroke 5 2 Cardiac related: Cardiogenic shock 1 0 Acute cardiac failure/ myocardial infarction Atrial or ventricular 3 2 fibrillation Cardiomyopathy 1 0 Others Unknown 24 17

4 Ann Thorac Surg KOERTKE ET AL 2007;83:24 9 INR SELF-MANAGEMENT AND LONG-TERM SURVIVAL Table 3. Results of Univariate Analysis for Long-Term Mortality Parameter Intent-to-Treat Analysis Exp (B) and 95% CI for Exp (B) N 930 Per Protocol Analysis Exp (B) and 95% CI for Exp (B) N CARDIOVASCULAR Age (years) 1.05 ( ) ( ) Study group (SEM group) 0.62 ( ) ( ) Atrial fibrillation (yes) 1.35 ( ) ( ) Other than aortic surgery 1.36 ( ) ( ) CI confidence interval; Exp(B) exponent of the B coefficient; SEM self-management. insufficiency was higher and the percentage of patients with an unknown primary diagnosis was lower in the SEM group compared with the CON group. Moreover, the SEM group received slightly more aortic valve prostheses and less double valve prostheses than the CON group. In addition, more St. Jude Medical valves (St. Jude Medical, Inc, St. Paul, MN) and less Medtronic Hall valves (Medtronic, Minneapolis, MN) were implanted in the SEM group compared with the CON group. Other parameters (such as body mass index, left ventricular ejection fraction, and the percentage of patients who were smokers or had hypertension or atrial fibrillation) did not differ between study groups. The distribution of the INR values in the CON group and the SEM group within the first two years of the ESCAT I study is given in Figure 1. In the CON group, a higher percentage of INR values were lying below the INR target range compared with the SEM group. Moreover, INR values were significantly higher in the SEM group compared with the CON group ( vs ; p 0.001). The average follow-up period of the 930 participants of the follow-up study was years. In total, the 930 patients (442 from the CON group and 488 from the SEM group) accrued 8,315 years of observation. One hundred twenty-two patients were either noncompliant within the first two years of the study or switched over from INR self-management to INR measurement by a general practitioner and vice versa during follow-up. These patients were excluded from the per protocol data analysis. In the patients who switched over from the INR measurement by a general practitioner to INR selfmanagement (N 50), age, sex distribution, percentage of hypertensive patients, and percentage of smokers was similar compared with the patients who switched over from INR self-management to INR management by a general practitioner (N 27) (p values 0.05, data not shown). During follow-up, 236 patients died. In the CON group, 1-year, 2-year, 5-year, and 10-year survival was 95.8%, 94.1%, 86.1%, and 76.3%, respectively. The corresponding values for the SEM group were 98.8%, 97.1%, 92.8%, and 84.5%. Thus, overall mortality was significantly higher in the CON group compared with the SEM group (Fig 2). According to the per protocol analysis, 10-year survival was 67.7% in the CON group and 80.6% in the SEM group. Causes of death are given in Table 2. In both study groups the majority of patients died because of cardiacrelated deaths. However, the total number of cardiacrelated deaths was nearly twice as high in the CON group as in the SEM group. Although the total number of bleeding-related causes of deaths was low in both study groups, slightly more such evident bleeding-related deaths occurred in the SEM group compared with the CON group. Parameters that were predictors of longterm mortality in the univariate analysis are given in Table 3 according to both the intent-to-treat analysis and the per protocol analysis. In the multivariate analysis, only age, valve position, and type of INR management remained independent predictors of mortality (Table 4). Mortality risk increased by 4% to 5% for each year of age and by 31% to 54% for mitral and double valve position. In contrast, self-management reduced mortality rate by approximately 23% according to the intent-to-treat analysis and by 33% if the per protocol analysis was used (Table 4). Comment In this study, we could demonstrate that INR selfmanagement is associated with a significant reduction in long-term mortality compared with INR measurement by Table 4. Results of Multivariate Analysis for Long-Term Mortality Parameter Intent-to-Treat Analysis Exp (B) 95% and CI of Exp (B) N 930 Per Protocol Analysis Exp (B) 95% and CI of Exp (B) N 808 Age (years) 1.04 ( ) ( ) Other than aortic valve surgery 1.31 ( ) ( ) Study group (SEM group) 0.77 ( ) ( ) 0.01 CI confidence interval; Exp(B) exponent of the B-coefficient; SEM self-management.

5 CARDIOVASCULAR 28 KOERTKE ET AL Ann Thorac Surg INR SELF-MANAGEMENT AND LONG-TERM SURVIVAL 2007;83:24 9 the general practitioner. Other predictors of survival were just age and the position of the implanted valve. It is already well-known that older age and mitral or double valve instead of aortic valve replacement are predictors of worse survival [11, 12]. Although the number of evident anticoagulationrelated deaths was small in both study groups, we cannot definitively rule out that our study design has underestimated anticoagulation-related causes of death. It is noteworthy that no autopsies were performed in the patients who died during follow-up. Table 2 demonstrates that especially the number of cardiac-related deaths was much higher in the CON group than in the SEM group. It is known that thromboembolism can be a cause of myocardial infarction [13]. Moreover, atrial and ventricular fibrillation, which can both lead to cardiac arrest, can be caused by coronary thromboembolism. The higher percentage of INR values below the target range, as well as the lower mean INR value in the CON group compared with the SEM group, can increase the risk of thromboembolic events. Recently [10], we have demonstrated that thromboembolic events occurred more often in the CON group than in the SEM group of our study, at least during the first two years of mechanical valve implantation. Thus, the higher number of cardiac-related deaths in the CON group compared with the SEM group can be reasonably related to the conventional strategy of oral anticoagulation. This assumption is also supported by the fact that per protocol analysis instead of the intentto-treat data analysis resulted in a more pronounced improvement of long-term survival in the SEM group compared with the CON group. Only two deaths in the CON group but five deaths in the SEM group were evidently related to bleeding events. Currently, there is an ongoing discussion concerning the INR target range that should be recommended in order to minimize anticoagulation-related complications. A recently performed meta-analysis indicates that the total number of thromboembolic and bleeding events was lower in the group with INR values above 3.0 compared with low-dose INR therapy ( 3.0) for both aortic and mitral valve prostheses [14]. However, others [15] have recommended target ranges of only 2.0 to 3.0 INR for aortic valve prostheses and 2.5 to 3.5 INR for mitral valve prostheses. In line with this lower target range, it has been demonstrated that a minimum risk of death was attained at 2.3 INR for patients with mechanical heart valve prostheses. With an increase of one unit of INR above 2.5, the risk of death from cerebral bleeding and from any cause was nearly doubled [16]. Because the mean INR values were significantly higher in the SEM group than in the CON group at least within the first two study years, we cannot definitively rule out that some of the bleedingrelated deaths in the SEM group were indeed related to higher INR values in this group compared with the CON group. In the ongoing ESCAT II trial, one group performed INR self-management with a target INR range of 1.8 to 2.8 for aortic valve replacement and 2.5 to 3.5 for mitral or double valve replacement, while another group performed INR self-management with a target INR range of 2.5 to 4.5 for all heart valve prostheses. In an interim analysis, total complication rate of grade III bleeding and thromboembolic events tended to be lower in the low-dose group than in the conventional-dose group [17]. Future analyses must show whether or not low-dose INR self-management is able to further increase long-term survival compared with conventional-dose INR self-management after mechanical heart valve replacement. The implantation of mechanical prosthetic valves is currently recommended, especially for patients below the age of 60 years, while patients after the age of 70 years should be provided with a bioprosthesis. At present, there is no consensus as to the choice of prosthetic heart valve for patients in their seventh decade of life [18]. Mean age of our study population at follow-up was 70 years. It has recently been demonstrated that the incidence of anticoagulation-related complications such as bleeding and thromboembolic events increase sharply with advanced age [19]. Because oral anticoagulation is required lifelong, some may argue that implantation of bioprostheses should be preferred instead of mechanical prostheses in order to reduce the risk of anticoagulation-related complications with advanced age. However, in an interim analysis in 1,818 patients of the ESCAT II trial with an age range of 18 to 82 years, we observed no age-related differences in the percentage of INR values lying in the target range [17]. Although some elderly patients and patients with difficulty in self-management of anticoagulation may need professional help in controlling anticoagulant dosage, INR self-management can be an alternative to conventional dose INR self-management in order to reduce anticoagulation-related complications in elderly patients. Moreover, the increased risk of thromboembolism with advanced age may at least in part be related to low fluid intake [20]. Consequently, drinking behavior should be improved in elderly patients by selective educational measures. In addition, bioprostheses bear the risk of valve-related complications that make a reoperation of the bioprosthesis necessary. Thus, implantation of mechanical valves may not be a relative contraindication in elderly people, but may be an alternative to the implantation of a bioprosthesis [21]. Our data of an improved survival in patients with INR self-management are in line with results of low morbidity and mortality in other patient groups who underwent effective patient training as an integral part of their therapy [1, 2]. Results indicate that education, as well as self-management of medical therapy, can improve the outcome of patients who require long-term treatment. In summary, our data demonstrate that INR selfmanagement is a promising strategy in order to increase long-term survival in patients with mechanical prosthetic valves.

6 Ann Thorac Surg KOERTKE ET AL 2007;83:24 9 INR SELF-MANAGEMENT AND LONG-TERM SURVIVAL References 1. Mühlhauser I, Berger M. Diabetes education and insulin therapy. When will they ever learn? J Intern Med 1993; 233: Sawicki PT, Muhlhauser I, Didjurgeit U, Baumgartner A, Bender R, Berger M. Intensified antihypertensive therapy is associated with improved survival in type I diabetic patients with nephropathy. J Hypertens 1995;13: Koertke H, Zittermann A, Mommertz S, El-Arousy M, Litmatzhe J, Koerfer R. The Bad Oeynhausen concept of INR self-management. J Thromb Thrombolysis 2005;19: White RH, McCurdy SA, von Marendorff H, Woodruff DE, Leftgoff L. Home prothrombin time monitoring after the initiation of warfarin therapy. Ann Int Med 1989;111: Sawicki P. A structured teaching and self-management program for patients receiving oral anticoagulation. JAMA 1999; 281: Cromheecke ME, Levi M, Colly LP, et al. Oral anticoagulation self-management and management by a specialist anticoagulation clinic: a randomised cross-over comparison. Lancet 2000;356: Watzke HH, Forberg E, Svolba G, Jimenez-Boj E, Krinniger B. A prospective controlled trial comparing weekly self-testing and self-dosing with the standard management of patients on stable oral anticoagulation. Thromb Haemost 2000;83: Christensen TD, Andersen NT, Attermann J, Hjortdal VE, Maegaard M, Hasenkam JM. Mechanical heart valve patients can manage oral anticoagulation therapy themselves. Eur J Cardiothorac Surg 2003;23: Körtke H, Körfer R. International normalized ratio selfmanagement after mechanical heart valve replacement: is an early start advantageous? Ann Thorac Surg 2001;72: Körtke H, Minami K, Breymann T, et al. INR self-management after mechanical heart valve replacement: ESCAT (Early Self- Controlled Anticoagulation Trial). Z Kardiol 2001;90(suppl 6): Cen YY, Glower DD, Landolfo K, et al. Comparison of survival after mitral valve replacement with biologic and mechanical valves in 1139 patients. J Thorac Cardiovasc Surg 2001;122: Parker DJ. Long-term morbidity and mortality after aortic and mitral valve replacement with tissue valves and certification to fly. Eur Heart J 1988;9(suppl G): Chesebro JH, Fuster V. Antithrombotic therapy for acute myocardial infarction: mechanisms and prevention of deep venous, left ventricular, and coronary artery thromboembolism. Circulation 1986;74:III Vink R, Kraaijenhagen RA, Hutten BA, et al. The optimal intensity of vitamin K antagonists in patients with mechanical heart valves: a meta-analysis. J Am Coll Cardiol 2003;42: Stein PD, Alpert JS, Bussey HI, Pharm D, Dalen JE, Turpie GG. Antithrombotic therapy in patients with mechanical and biological prosthetic heart valves. Chest 2001;119(suppl 1):220S Oden A, Fahlen M. Oral anticoagulation and risk of death: a medical record linkage study. Br Med J 2002;325: Koertke H, Zittermann A, Minami K, et al. Low-dose international normalized ratio self-management: a promising tool to achieve low complication rates after mechanical heart valve replacement. Ann Thorac Surg 2005;79: Borel ML, Amaral A, Bezon E, et al. Evaluation de la mortalite et de la qualite de vie postoperatoire des sujets de plus de 75 ans apres remplacement valvulaire pour retrecissement aortique. A propos de 110 cas. Arch Mal Coeur Vaiss 2003;96: Torn M, Bollen WLEM, van der Meer FJM, van der Wall EE, Rosendaal FR. Risks of oral anticoagulation therapy with increasing age. Arch Intern Med 2005;165: Minami K, Zittermann A, Schulte-Eistrup S, Koertke K, Körfer R. Mitroflow synergy prostheses for aortic valve replacement: 19 years experience with 1,516 patients. Ann Thorac Surg 2005;80: Eitz T, Fritzsche D, Kleikamp G, Zittermann A, Horstkotte D, Körfer R. Reoperation of the aortic valve in octogenarians. Ann Thorac Surg 2006;82: CARDIOVASCULAR