New oral anticoagulants show favorable safety, efficacy vs. warfarin

Richard C. Becker, MD, FAHA

Collectively, this is very compelling data that speaks to the  benefit of a new generation of oral anticoagulants compared with warfarin.

The major takeaway points for me were as follows: First, the hemorrhagic stroke rate being reduced by 50% is very important, because this type of stroke carries a high mortality and morbidity. We’ve talked for years about the potential effect of warfarin on the blood vessels themselves; there’s a relatively old term, “warfarin vasculopathy,” that proposes that warfarin exerts a detrimental effect on the integrity of  blood vessels, particularly in the central nervous system. This may be related directly to vitamin K-dependent proteins within the vessel wall. I believe the observations surrounding hemorrhagic stroke  support  this line of thinking.

The second important takeaway was that the overall effects were neutral for stroke and systemic embolism at the lower doses of the new oral anticoagulants. There was a 28% increase in ischemic stroke, an increase in MI, but lower bleeding risk and reduced mortality compared with warfarin. 

So what does that mean? From my perspective, the use of lower doses (specifically for dabigatran and edoxaban) requires careful thought among clinicians, with inclusion of the patient’s values and preferences. Those at relatively low risk for stroke but with high bleeding risk may be well served by lower doses. I will also say that additional investigation is needed to complement known pharmacokinetic and pharmacodynamic parameters and patient characteristics that factor into the overall clinical equation.

The last takeaway was the increased likelihood of major GI bleeding that has been observed consistently with the new oral anticoagulants compared to warfarin. We still lack a satisfying explanation as to why this is the case. Clearly, further investigation to include an understanding of drug effects at the brush border of the intestine will be needed.

Albert L. Waldo, MD

This generally well-done meta-analysis confirms what we knew individually about the new oral anticoagulants (NOACs) versus warfarin. As discussed thoroughly in the paper, the NOACs have several important advantages over warfarin, including that they have minimal interactions with food and drugs, and rapid onset of action. This meta-analysis appropriately emphasizes the NOACs’ significantly less intracranial hemorrhage and hemorrhagic stroke compared to warfarin.

While the latter are important and impressive, all the data were discussed as percentage differences. The absolute differences were not considered, and I was a little troubled by that, because talking only  about percentage difference masks the actually very small absolute differences between the NOACs and warfarin in the important outcomes of intracranial hemorrhage and hemorrhagic stroke: 0.44% per year for intracranial bleeds, and 0.19%/year for hemorrhagic stroke.

While these differences are statistically significant and important, they should be considered in perspective. For instance, there may be a consideration whether to switch from warfarin to one of the NOACs, in which circumstance, it is not clear that patients with well-managed warfarin should be switched to one of the NOACs. The absence of an antidote for the NOACs is another such consideration. Also, warfarin is significantly less expensive. Thus, while the NOACs are, in general, a big advance, it’s not always a slam-dunk that you should stop using warfarin.

Also, in this meta-analysis, some of the differences between the various drugs are lost. For instance: rivaroxaban, dabigatran, and high-dose edoxaban increased GI bleeding compared to warfarin, but apixaban and low-dose edoxaban decreased GI bleeding compared to warfarin.

It is important to note that there are several categories of patients who were not covered in the studies that were a part of the meta-analysis. One – and this is a big concern – is the huge number of patients who have both CAD and atrial fibrillation. We know that the combination of anti-platelet agents and oral anticoagulants increases bleeding significantly, and if you have to administer triple therapy (an oral anticoagulant with two anti-platelet agents), about one in four patients will have significant bleeding. 

In patients with CAD who have AF and are at risk for stroke, we clearly need some well-designed, prospective studies to guide us on what to do. A small, prospective study, the WOEST trial, indicated that double therapy with warfarin plus clopidogrel was as effective as triple therapy (the latter two plus aspirin). Also, a large, retrospective Danish study came to the same conclusion. In short, there are precious few data on which to base the clinical approach to these patients, leaving a very big need.