Author + information
Vink et al. (1) compared the occurrence rates of bleeding and thrombo-embolism among patients with mechanical heart valves who had either high-intensity or low-intensity vitamin K antagonist therapy (VKA). Based on their findings of lower combined bleeding and thrombo-embolic occurrence rates in the high- versus low-intensity VKA group, they concluded that “both aortic and mitral valves will benefit from a treatment strategy with a target INR higher than 3.0.” This is a controversial conclusion because most recent publications recommend a valve-specific and lower target international normalized ratio (INR) (2,3). Therefore, we closely studied Vink's analyses and have the following comments.
When trying to identify the optimal target INR for patients with mechanical heart valves, not only the occurrence rates of bleeding and thrombo-embolism but also the consequences of these events need to be considered. It may well be possible that although the combined occurrence rates of bleeding and thrombo-embolism are lower in the high-intensity VKA group, the combined mortality associated with bleeding and thromboembolism is lower in the low-intensity VKA group.
We repeated the reported meta-analysis for the aortic valve replacement group, and we pooled the mortality rates of bleeding and thrombo-embolism from those studies that reported on mortality resulting from these events (11 of the 21 studies in the low-intensity and 5 of the 9 studies in the high-intensity VKA group). We found in the low-intensity VKA group versus the high-intensity VKA group, a pooled mortality rate for bleeding of 12% (95% confidence interval [CI] 7.7% to 16.3%) versus 20% (95% CI 13.3% to 26.6%; p = 0.05), for valve thrombosis 27% (95% CI 12% to 42%) versus 33% (95% CI 3.5% to 63.5% p = NS), and for thrombo-embolism 14% (95% CI 9.6% to 18.4%) versus 14% (95% CI 8.3% to 19.7% p = NS). We entered the occurrence rates and mortality rates of bleeding, valve thrombosis, and thrombo-embolism from the meta-analysis in a microsimulation model that we previously developed (4) to estimate patient prognosis after aortic valve replacement with mechanical prostheses, and we simulated 10,000 life histories of a 56-year-old (mean age in meta-analysis) male patient after aortic valve replacement in order to estimate the expected lifetime risk of death due to bleeding, valve thrombosis, and thrombo-embolism for either a low-intensity or high-intensity anticoagulation regimen. We found that the combined lifetime risk of death due to bleeding, valve thrombosis, and thrombo-embolism in the low-intensity versus the high-intensity VKA group was 6.3% versus 7.8% (p = NS). This shows that, when considering the mortality due to bleeding and thrombo-embolic events, for aortic valve patients nothing is gained by a target INR higher than 3.0.
In conclusion, a meta-analysis of only bleeding and thrombo-embolic event rates is methodologically insufficient to determine the optimal anticoagulation regimen for patients with mechanical heart valves. Modelling techniques are essential to evaluate the burden of the competing risks of bleeding and thrombo-embolism and their consequences during the lifetime of the patient.
- American College of Cardiology Foundation
- Vink R.,
- Kraaijenhagen R.A.,
- Hutten B.A,
- et al.
- ↵Huth C, Friedl A, Rost A. Intensity of oral anticoagulation after implantation of St. Jude Medical aortic prosthesis: analysis of the GELIA database (GELIA 4). Eur Heart J Suppl 2001;Suppl Q:Q33–8.
- ACC/AHA guidelines for the management of patients with valvular heart disease. A report of the American College of Cardiology/American Heart Association. Task Force on Practice Guidelines (Committee on Management of Patients with Valvular Heart Disease). J Am Coll Cardiol 1998;32:1486–588.