Abstract
Purpose: Early antithrombotic therapy after bioprosthetic aortic valve replacement (AVR) is controversial. This study aimed to retrospectively compare between warfarin and aspirin treatment in the 3 months after bioprosthetic AVR for elderly patients more than 60 years old, and to determine the optimal antithrombotic therapy.
Methods: This retrospective study included 479 patients in single center from January 1994 to June 2014. Patients were divided into two groups (Wa group, warfarin; As group, aspirin). We searched our computerized clinical database for thromboembolic or bleeding events. Propensity score analysis was conducted to adjust for selection bias.
Results: All patients, except one patient, were followed-up in the out-patient department for 3 months after the operation. In all, 86 propensity-matched patient-pairs were derived. Early operative outcomes were similar in both the groups. There are one patient of thromboembolic event and three patients of bleeding events, but the prevalence was not significantly different (p >0.999).
Conclusion: The incidence of thromboembolic and bleeding events during early 3 months after bioprosthetic AVR were similar in Wa and As groups. If the patient does not have indications of warfarin, early antithrombotic therapy with aspirin only may be easier and more feasible for elderly patients.
Keywords: aortic valve replacement, bioprostheses, oral anticoagulation
Introduction
Aortic valve replacement (AVR) is one of the most commonly performed treatments for aortic valve disease. Bioprostheses have less durability than mechanical valves, but one advantage over mechanical prostheses is freedom from anticoagulation. In many centers, bioprostheses are chosen for elderly patients or patients who have risks or contraindications of warfarin.
Antithrombotic therapy after bioprosthetic AVR was controversial. Although many authors recommend vitamin K antagonist treatment in the first 90 days after valve replacement until the prosthetic valve is fully endothelialized, recent guidelines for bioprosthetic valve replacement recommend aspirin for patients in sinus rhythm without indications of warfarin.1–3) Most studies to support the current guidelines of early aspirin anticoagulation were based on B or C levels of evidence.
The purpose of this study was to retrospectively compare warfarin and aspirin treatment in the 3 months after bioprosthetic AVR in elderly patients more than 60 years old and to evaluate the benefits from thromboembolic or bleeding complications.
Materials and Methods
Patients
From January 1994 to June 2014, 479 patients over 60 years of age underwent bioprosthetic AVR in our institution. Cases were identified through a search of our computerized clinical database, retrospectively. Patients were divided into two groups. One group was received warfarin (Wa group, n = 312) and the other group was treated with aspirin (As group, n = 167).
Most patients underwent an operation for aortic valve stenosis or regurgitation. Patients with coronary disease or aortic dilation underwent coronary artery bypass grafting (CABG), aortic replacement, or aortoplasty. Maze operation was performed for patients with atrial fibrillation. We included patients with other valvular diseases that were not needed antithrombotic therapy such as mitral valvuloplasty and tricuspid valvuloplasty without ring annuloplasty.
The study was approved by the institutional ethics committee/review board of the Asan Medical Center, and the requirement for informed patient consent was waived in view of the retrospective nature of the study.
Medication
Antithrombotic medication was begun as soon as the patient resumed oral intake. When extubation of endotracheal tube was delayed until more than 48 hours later, the patient was taken a medication through nasogastric tube. The type of antithrombotic therapy was determined by the surgeon’s preference, but the patients who have the indications of warfarin, such as postoperative atrial fibrillation, were received a warfarin. Low molecular-weight heparin had been injected in the warfarin group until an international normalized ratio (INR) of 1.5 was achieved, and the dose of warfarin was modified to maintain an INR between 1.5 and 2.5 in the out-patient department. The As group received antiplatelet therapy with low-dose aspirin (100 mg/daily).
Follow-up
All patients were followed-up in our institute during 3 months after the operation. We searched our computerized clinical database for thromboembolic or bleeding events. In out-patient department, we checked the patient’s symptoms, particularly in terms of aphasia, weakness, mental changes, melena, hematochezia, or hematuria. All patients who had suspected thromboembolic or bleeding events underwent computed tomography (CT), magnetic resonance imaging (MRI), or endoscopy and consulted physicians of neurology or internal medicine departments. Asymptomatic patients with asymptomatic peripheral embolic events could not be identified. The thromboembolic or bleeding events were defined as morbidities confirmed by imaging studies or consultation with the specialist.
Statistical analysis
Data with continuous variables were expressed as means ± standard deviation and categorical variables were expressed as numbers and percentages. Continuous variables were compared using Student’s t-test. Categorical variables were compared using the χ2 test or Fisher’s exact test as appropriate.
To reduce the effect of selection bias, we conducted the analysis using propensity score matching. The propensity score, which is the probability that a patient would be selected for As group based on the measured baseline variables, was calculated by multiple logistic regression analysis based on preoperative covariates shown in Table 1 as well as operative profiles listed in Table 2, and was used to match Wa group with As group (1:1 match). As a threshold for statistically significant differences, p value <0.05 was used. SPSS version 21.0 (Korean version; IBM Corporation, USA) was used for statistical analyses.
Table 1. Characteristics of the patients and preoperative echocardiographic data.
| Wa | As | p-value | |
|---|---|---|---|
| Number of patients | 312 | 167 | |
| Male gender, n (%) | 163 (52.2%) | 91 (54.5%) | 0.701 |
| Age, year ± SD | 71.43 ± 4.80 | 71.25 ± 6.12 | 0.749 |
| Body surface area, cm2 ± SD | 1.59 ± 0.17 | 1.60 ± 0.16 | 0.638 |
| Body mass index, kg/m2 ± SD | 25.38 ± 2.44 | 23.87 ± 3.58 | 0.433 |
| Diabetes mellitus, n (%) | 60 (19.2%) | 37 (22.2%) | 0.448 |
| Hypertension, n (%) | 147 (47.1%) | 72 (43.1%) | 0.402 |
| Cerebrovascular accident, n (%) | 20 (6.4%) | 17 (10.2%) | 0.141 |
| COPD, n (%) | 17 (5.4%) | 5 (3.0%) | 0.259 |
| Chronic kidney injury, n (%) | 6 (1.9%) | 4 (2.4%) | 0.731 |
| Coronary artery disease | 10 (3.2%) | 56 (33.5%) | <0.001 |
| Atrial fibrillation | 12 (3.8%) | 1 (0.6%) | 0.037 |
| NYHA functional class, n (%) | 0.039 | ||
| I | 70 (22.4%) | 55 (32.9%) | |
| II | 147 (47.1%) | 77 (46.1%) | |
| III | 67 (21.5%) | 26 (15.6%) | |
| IV | 28 (9.0%) | 9 (5.4%) | |
| Aortic regurgitation ≥III, n (%) | 112 (35.9%) | 48 (28.7%) | 0.114 |
| Severe aortic stenosis, n (%) | 253 (81.1%) | 132 (79.0%) | 0.591 |
| Preoperative LVEF, % ± SD | 56.34 ± 12.92 | 57.15 ± 12.37 | 0.506 |
| LVEF <30, n (%) | 23 (7.4%) | 9 (5.4%) | 0.408 |
| Pre-LVIDs, mm ± SD | 35.37 ± 10.63 | 34.37 ± 10.09 | 0.391 |
| Pre-LVIDd, mm, ± SD | 52.71 ± 9.84 | 52.08 ± 10.32 | 0.509 |
COPD: chronic obstructive pulmonary disease; NYHA: New York Heart Association; LVEF: left ventricular ejection fraction; pre-LVIDs: preoperative left ventricular internal dimension in the systolic phase; pre-LVIDd: preoperative left ventricular internal dimension in the diastolic phase; SD: standard deviation; Wa group: warfarin; As group: aspirin
Table 2. Operative data and early complications.
| Wa (312) | As (167) | p-value | |
|---|---|---|---|
| CPB time, min ± SD | 116.8 ± 42.0 | 125.2 ± 49.7 | 0.069 |
| ACC time, min ± SD | 74.8 ± 26.7 | 76.7 ± 28.1 | 0.487 |
| Bovine, n (%) | 215 (68.9%) | 113 (67.7%) | 0.724 |
| Valve size, mm ± SD | 21.80 ± 2.19 | 23.15 ± 2.28 | <0.001 |
| Cooperation, n (%) | |||
| CABG | 10 (3.2%) | 56 (33.5%) | <0.001 |
| Mitral valvuloplasty | 1 (0.3%) | 2 (1.2%) | 0.246 |
| Tricuspid valvuloplasty | 4 (1.3%) | 2 (1.2%) | 0.937 |
| Aorta replacement or aortoplasty | 39 (12.5%) | 22 (13.2%) | 0.833 |
| Maze procedure | 6 (1.9%) | 0 (0.0%) | 0.071 |
| Early complications, n (%) | |||
| Postoperative bleeding | 16 (5.1%) | 8 (4.8%) | 0.872 |
| Low cardiac output syndrome | 1 (0.3%) | 1 (0.6%) | 0.653 |
| Pleural effusion | 1 (0.3%) | 2 (1.2%) | 0.246 |
| Acute kidney injury | 2 (0.6%) | 4 (2.4%) | 0.100 |
| Wound infection | 1 (0.3%) | 5 (3.0%) | 0.012 |
| Cerebral infarction | 1 (0.3%) | 2 (1.2%) | 0.246 |
ACC: aortic cross clamp; CABG: coronary artery bypass grafting surgery; CPB: cardiopulmonary bypass; SD: standard deviation; Wa group: warfarin; As group: aspirin
Results
The preoperative patient characteristics of two groups are summarized in Table 1. All patients, except for one patient, were followed-up in the out-patient department during the 3 months after the operation. One patient was admitted to hospital over the 3 months. The prevalence of coronary artery disease and atrial fibrillation was significantly different between two groups.
Table 2 shows the operative data and early complications. The valve size of patients in As group was larger than in Wa group. Bovine valves were used more frequently than porcine valves in all groups.
After propensity score matching, 86 pairs of patients were matched and no significant difference between the two groups was observed (Table 3). The distribution of propensity score (PS) before and after matching was demonstrated in Fig. 1.
Table 3. Baseline features and postoperative outcomes of propensity score matched patients.
| Wa | As | p-value | |
|---|---|---|---|
| Number of patients | 86 | 86 | |
| Male gender, n (%) | 45 (52.3%) | 43 (50.0%) | 0.760 |
| Age, year ± SD | 70.98 ± 4.72 | 71.62 ± 6.12 | 0.444 |
| Body surface area, cm2 ± SD | 1.58 ± 0.17 | 1.60 ± 0.17 | 0.416 |
| Body mass index, kg/m2 ± SD | 23.30 ± 3.11 | 23.87 ± 3.83 | 0.288 |
| Diabetes mellitus, n (%) | 17 (19.8%) | 16 (18.6%) | 0.846 |
| Hypertension, n (%) | 37 (43.0%) | 38 (44.2%) | 0.878 |
| Cerebrovascular accident, n (%) | 3 (3.5%) | 8 (9.3%) | 0.119 |
| COPD, n (%) | 2 (2.3%) | 3 (3.5%) | >0.999 |
| Chronic kidney injury, n (%) | 1 (1.2%) | 2 (2.3%) | >0.999 |
| Coronary artery disease | 7 (8.1%) | 7 (8.1%) | >0.999 |
| Atrial fibrillation | 1 (1.2%) | 1 (1.2%) | >0.999 |
| NYHA functional class, n (%) | 0.979 | ||
| I | 24 (27.9%) | 24 (27.9%) | |
| II | 40 (46.5%) | 42 (48.8%) | |
| III | 17 (19.8%) | 16 (18.6%) | |
| IV | 5 (5.8%) | 4 (4.7%) | |
| Aortic regurgitation ≥III, n (%) | 56 (65.1%) | 59 (68.6%) | 0.627 |
| Severe aortic stenosis, n (%) | 72 (83.7%) | 72 (83.7%) | >0.999 |
| Preoperative LVEF, % ± SD | 58.80 ± 12.30 | 57.40 ± 12.18 | 0.453 |
| LVEF <30, n (%) | 6 (7.0%) | 4 (4.7%) | 0.746 |
| Pre-LVIDs, mm ± SD | 33.90 ± 9.66 | 34.16 ± 10.17 | 0.860 |
| CPB time, min ± SD | 116.0 ± 44.3 | 118.2 ± 47.5 | 0.754 |
| ACC time, min ± SD | 74.3 ± 27.1 | 75.6 ± 28.1 | 0.772 |
| Bovine, n (%) | 57 (66.3%) | 58 (67.4%) | 0.871 |
| Valve size, mm ± SD | 22.66 ± 2.11 | 22.53 ± 2.07 | 0.689 |
| Cooperation, n (%) | |||
| CABG | 7 (8.1%) | 7 (8.1%) | >0.999 |
| Mitral valvuloplasty | 1 (1.2%) | 1 (1.2%) | >0.999 |
| Tricuspid valvuloplasty | 0 (0.0%) | 0 (0.0%) | |
| Aorta replacement or aortoplasty | 10 (11.6%) | 11 (12.8%) | 0.816 |
| Maze procedure | 0 (0.0%) | 0 (0.0%) | |
| Early complications related the surgery, n (%) | |||
| Postoperative bleeding | 2 (2.3%) | 5 (5.8%) | 0.443 |
| Low cardiac output syndrome | 0 (0.0%) | 0 (0.0%) | |
| Pleural effusion | 0 (0.0%) | 2 (1.2%) | 0.497 |
| Acute kidney injury | 0 (0.0%) | 2 (2.3%) | 0.497 |
| Wound infection | 1 (1.2%) | 4 (4.7%) | 0.368 |
| Cerebral infarction | 0 (0.0%) | 1 (1.2%) | >0.999 |
| Thromboembolic event during 90 days, n (%) | 1 (1.2%) | 0 (0%) | >0.999 |
| Cerebral infarction | 1 | ||
| Bleeding event during 90 days, n (%) | 2 (2.3%) | 1 (1.2%) | >0.999 |
| Upper gastrointestinal bleeding | 2 | 1 | |
ACC: aortic cross clamp; CABG: coronary artery bypass grafting surgery; CPB: cardiopulmonary bypass; COPD: chronic obstructive pulmonary disease; NYHA: New York Heart Association; LVEF: left ventricular ejection fraction; pre-LVIDs: preoperative left ventricular internal dimension in the systolic phase; pre-LVIDd: preoperative left ventricular internal dimension in the diastolic phase; SD: standard deviation; Wa group: warfarin; As group: aspirin
Fig. 1. Distribution of variable after propensity score matching.
The thromboembolic or bleeding event during first 90 days was identified in four patients (Table 3). One patient suffered cerebral infarction with global aphasia at 1 month after surgery and confirmed by brain MRI.
There are three patients who confirmed bleeding complication. One patient had hematochezia on postoperative day 43 and the others complained melena within 2 weeks after surgery (Fig. 2). They were diagnosed with gastric ulcer bleeding using esophagogastroduodenoscopy. Three patients were already taken anti-ulcer medication with H2 receptor blocker when they were confirmed with ulcer bleeding. The INR of two patients in Wa group were 2.92 and 10.62, respectively. The age of two patients is over 70 years old. The prevalence of thromboembolic and bleeding event was not significantly different.
Fig. 2. Timeline of thromboembolic and bleeding events.
Discussion
We compared clinical outcomes of two different antithrombotic regimens through propensity score matching analysis to reduce a bias. Our study is first evaluation of antithrombotic regimens after bioprosthetic AVR in Asian population. Although we expected significant difference in thromboembolic or bleeding events, there were no benefits or risks of specific antithrombotic therapy. Similar analysis for American was performed previously.4) The authors demonstrated that prevalence of early embolic or bleeding events after bioprosthetic AVR is low. We found similar trend prevalence of thromboembolic and bleeding events. Although they concluded that aspirin plus warfarin was associated with a reduced risk of death and embolic events, we could not find the similar results. It caused from low incidence of thromboembolic and bleeding events. Brennan et al.4) also mentioned that aspirin plus warfarin was associated with an increased bleeding risk.
Early anticoagulation with warfarin may help to decrease the risk of thromboembolism until the prosthetic valve is fully endothelialized. The most guidelines recommend early anticoagulation after bioprosthetic AVR with warfarin during first 3 months. However, the present study shows that taking warfarin during the first 90 postoperative days did not have additional benefits for patients who underwent bioprosthetic AVR.5–8) The recent guidelines for anticoagulation of bioprosthetic aortic valves were changed from warfarin to aspirin for patients of low thromboembolic risk based on these backgrounds.1–3) Also, current bioprosthetic aortic valves have very different properties than the first-generation valves. New-generation valves are less thrombogenic and have a reduced incidence of structural failure.9,10)
It is more difficult to receive warfarin treatment in the elderly patients over 60 years old because aspirin whereas just taking a day, warfarin interacts with the foods, drugs and herbal medicines, especially in Korea. It takes discomfort of patients and social costs until the INR is stable. Warfarin therapy has a significant risk of morbidity with bleeding and limitations of daily life. Recently, the development of testing devices for INR self-management has enabled patients to better control warfarin treatment; however, this is still not easy for elderly patients. In this regard, the early antithrombotic therapy after bioprosthetic AVR was changed from warfarin to aspirin, and our results support that this paradigm is changed.
Several studies found old age and higher INR were risk factors of hemorrhage related to warfarin.11–13) Our patients suffered ulcer bleeding in Wa group were over 70 years old and the INR of one patient was 10.62. Although our results demonstrate a lack of apparent benefits of early aspirin therapy on bleeding, previous meta-analysis study comparing warfarin and aspirin suggests that frequency of bleeding in patients receiving warfarin was higher than aspirin.14) Aspirin, especially for elderly patients, may be the most appropriate early anticoagulation therapy after bioprosthetic AVR.
However, risk factors for thromboembolism include age, tobacco use, hypertension, diabetes, hyperlipidemia, atrial fibrillation, ventricular function, left atrial size (>50 mm on echocardiography), previous thromboembolism, and abnormalities of the coagulation system including hepatic failure.15) We also treated with a warfarin for patients with postoperative atrial fibrillation and huge left atrial size (>50 mm on echocardiography).
The limitations of our study include natures in retrospective studies of observational data from a single center. Although a propensity score matched analysis was used in this study, it still has some bias. Furthermore, our sample size remains too small to compared two antithrombotic regimens. Although the event rates differ, several reports also concluded with small event rate.6) Postoperative atrial fibrillation is very important factor related with systemic embolic event but we could not treat with an aspirin because of ethical issues. We followed the guideline of management for atrial fibrillation and all patients with postoperative atrial fibrillation were treated with a warfarin.
Conclusion
Early anticoagulation with warfarin does not have definite benefits of thromboembolism and risks of bleeding. If the patient does not have indication of warfarin, early antithrombotic therapy with aspirin may be easier and more feasible for elderly patients because INR control of warfarin for elderly patients is not convenient.
Acknowledgments
We thank the staff of the neurology and internal medicine department of our hospital for the diagnosis of thromboembolism and bleeding, and the Division of Biostatistics, Center for Medical Research and Information, University of Ulsan College of Medicine, for their assistance with the statistical analysis.
Disclosure Statement
The authors declare that they have no conflicts of interest.
References
- 1).Nishimura RA, Otto CM, Bonow RO, et al. 2014 AHA/ACC guideline for the management of patients with valvular heart disease: executive summary: a report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines. J AM Coll Cardiol 2014; 63: e57-185. [DOI] [PubMed] [Google Scholar]
- 2).Vahanian A, Alfieri O, Andreotti F, et al. Guidelines on the management of valvular heart disease (version 2012). Eur Heart J 2012; 33: 2451-96. [DOI] [PubMed] [Google Scholar]
- 3).Whitlock RP, Sun JC, Fremes SE, et al. Antithrombotic and thrombolytic therapy for valvular disease: antithrombotic therapy and prevention of thrombosis, 9th ed: American college of chest physicians evidence- based clinical practice guidelines. Chest 2012; 141: e576S-600S. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4).Brennan JM, Edwards FH, Zhao Y, et al. Early anticoagulation of bioprosthetic aortic valves in older patients: results from the Society of Thoracic Surgeons Adult Cardiac Surgery National Database. J Am Coll Cardiol 2012; 60: 971-7. [DOI] [PubMed] [Google Scholar]
- 5).Sundt TM, Zehr KJ, Dearani JA, et al. Is early anticoagulation with warfarin necessary after bioprosthetic aortic valve replacement? J Thorac Cardiovasc Surg 2005; 129: 1024-31. [DOI] [PubMed] [Google Scholar]
- 6).El-Husseiny M, Salhiyyah K, Raja SG, et al. Should warfarin be routinely prescribed for the first three months after a bioprosthetic valve replacement? Interact Cardiovasc Thorac Surg 2006; 5: 616-23. [DOI] [PubMed] [Google Scholar]
- 7).Gherli T, Colli A, Fragnito C, et al. Comparing warfarin with aspirin after biological aortic valve replacement: a prospective study. Circulation 2004; 110: 496-500. [DOI] [PubMed] [Google Scholar]
- 8).Mydin MI, Dimitrakakis G, Younis J, et al. Optimal thromboprophylaxis following bioprosthetic aortic valve replacement: still a matter of debate? Interact Cardiovasc Thorac Surg 2012; 15: 109-14. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9).Rahimtoola SH. Choice of prosthetic heart valve in adults an update. J Am Coll Cardiol 2010; 55: 2413-26. [DOI] [PubMed] [Google Scholar]
- 10).Minakata K, Tanaka S, Okawa Y, et al. Long-term outcome of the carpentier-edwards pericardial valve in the aortic position in Japanese patients. Circ J 2014; 78: 882-9. [DOI] [PubMed] [Google Scholar]
- 11).Levine MN, Raskob G, Beyth RJ, et al. Hemorrhagic complications of anticoagulant treatment: the Seventh ACCP Conference on Antithrombotic and Thrombolytic Therapy. Chest 2004; 126: 287S-310S. [DOI] [PubMed] [Google Scholar]
- 12).Hylek EM, Evans-Molina C, Shea C, et al. Major hemorrhage and tolerability of warfarin in the first year of therapy among elderly patients with atrial fibrillation. Circulation 2007; 115: 2689-96. [DOI] [PubMed] [Google Scholar]
- 13).Akhtar RP, Abid AR, Zafar H, et al. Aniticoagulation in patients following prosthetic heart valve replacement. Ann Thorac Cardiovasc Surg 2009; 15: 10-7. [PubMed] [Google Scholar]
- 14).Warkentin AE, Donadini MP, Spencer FA, et al. Bleeding risk in randomized controlled trials comparing warfarin and aspirin: a systematic review and meta-analysis. J Thromb Haemost 2012; 10: 512-20. [DOI] [PubMed] [Google Scholar]
- 15).Goldsmith I, Turpie AG, Lip GY. Valvular heart disease and prosthetic heart valves. BMJ 2002; 23: 325: 1228-31. [DOI] [PMC free article] [PubMed] [Google Scholar]