Warfarin vs Aspirin for Bioprosthetic Aortic Valve Replacement

Abstract & Commentary

Synopsis: Low-dose aspirin is as effective as warfarin for preventing cerebral thromboembolic events after bioprosthetic aortic valve replacement.

Source: Gherli T, et al. Circulation. 2004;110:496-500.

Current guidelines recommend warfarin anticoagulation for the first 3 months after biprosthetic aortic valve replacement. However, because of the risk of bleeding, some surgeons are comfortable with aspirin only during this period. Thus, Gherli and colleagues from Parma, Italy performed an observational study of patients undergoing bioprosthetic aortic valve replacement at 1 institution. Patients with atrial fibrillation at any time, multiple valve replacement, and any potential indication for warfarin therapy were excluded. Of the 9 senior surgeons, 5 gave aspirin and 4 gave warfarin, so patient assignment to therapy depended on who was operating the day of surgery. All patients got low molecular weight heparin on the first post-operative day, and 100 mg aspirin or warfarin was started on day 2. Those getting warfarin had the heparin continued until INRs between 2-3 were achieved. Warfarin was continued for 3 months, and then aspirin was substituted. In those with concomitant coronary artery bypass surgery, aspirin was withheld until warfarin was discontinued. The primary endpoints were cerebral ischemic events, bleeding, and survival. There were 3 cerebral ischemic events in the 141 patients (2.1%) receiving aspirin and 5 in the 108 patients (4.6%) on warfarin. After 3 months, and up to 16 months of follow-up, there was 1 additional cerebral ischemic event in the aspirin group (0.7%) and 3 in the warfarin group (2.8%).

There was no statistically, significant difference in cerebral ischemic events between the 2 groups. Major bleeding occurred in 3 patients in the aspirin group (2.1%), all were gastrointestinal bleeding, and 4 in the warfarin group (3.7%). In all cases of bleeding with warfarin, the INR was > 3.0 on readmission to the hospital. Perioperative survival was > 98% and long-term > 95%. There was no difference in survival or stroke-free survival between the 2 groups.

Gherli et al concluded that low-dose aspirin is as effective as warfarin for preventing cerebral thromboembolic events after bioprosthetic aortic valve replacement.

Comment by Michael H. Crawford, MD

Despite the recommendations by the ACC/AHA guidelines, there has been a move by cardiac surgeons away from 3 months of warfarin after bioprosthetic aortic valve replacement because of observational data and their own experience with the bleeding complications of warfarin, especially during the first 30 days post-operatively. Although this study did not show a statistically significant difference in bleeding complications, the type of bleeding with each therapy is instructive. All 3 patients in the aspirin group had gastrointestinal bleeding that abated with conservative therapy. The 4 episodes on warfarin were: 1 vein harvest site bleeding; 2 mediastinal bleeding; and 1 patient with an intracranial bleed who died. You can see why surgeons are not too keen on warfarin.

This prospective observational study may be biased since the referring physicians could have directed cases, where they preferred 1 therapy over another, to a day when the right surgeon was operating. The 2 groups were well matched, except that the warfarin group was slightly older (73 vs 70 years, P = .007) and their preoperative risk score was higher (6.9 vs 6.1; P = .015). It is difficult to know if this influenced the results. Of more concern is the fact that this was a low-risk group. Patients with any possible confounder were eliminated, including patients with peripheral vascular disease or previous anticoagulation therapy. The paper does not reveal how many patients were excluded.

Since there was no difference between the 2 groups in the major outcomes, one could establish a protocol or pick the therapy thought best for each patient. Unless there is another indication for warfarin such as atrial fibrillation, few would probably select warfarin given the inconvenience of taking it. However, when newer anticoagulants, that are less affected by food and other medications and do not require periodic blood tests, become available, the pendulum could swing the other way.

Dr. Crawford, Professor of Medicine, Associate Chief of Cardiology for Clinical Programs University of California San Francisco, is Editor of Clinical Cardiology Alert.