The Role of and Selection Process 
for the New Anticoagulants in 
Nonvalvular Atrial Fibrillation

Please subscribe to use our print features or to download PDF files.

Dr. Reiffel: Right, I agree with those comments. I also wrote in response to that FDA decision. We know in Europe it’s being used at the 110 mg dose, primarily as Peter identified in patients with modest renal insufficiency who are taking potentially drug interacting agents. And I must say, I’ve now seen probably about two dozen patients who have gotten the dose of 110 mg in another country and have come back with supplies. So there are some people using it. I’ve got physicians around me who referred in-patients  taking 75 mg three times daily or alternating 75/150 mg trying to mimic the daily dose of the 110 mg. So I would agree with Peter. I think this decision would have been better left to the clinicians than having taken the choice out of our hands by the FDA.

Dr. Naccarelli: In addition, you have the 150 mg dose being used in people with creatinine clearances of 32 on amiodarone and they had a bleed. Then it comes back to haunt the drug— “Oh, well your drug causes bleeding”—and I think if we had the 110 mg dose, people would have used that. Theoretically or based on the trials there was less bleeding on that dose.

Dr. Halperin: I think that was borne out by John Eikelboom’s paper based on a secondary analysis12 looking at the patients over 75, in whom there was an excess of major extracranial bleeding with 150 mg of dabigatran twice daily that did not appear in those given the 110 mg twice daily dose.

Dr. Reiffel: Correct. So I think this is a deficit in what we have so far. Jerry, could you make a couple of brief comments about the information we have regarding cardioversion with these new agents?

Dr. Naccarelli: As you know, one of the issues is that the embolic rate, if the patient has persistent atrial fibrillation and has been in atrial fibrillation more than two days, patients who have cardioversion, whether it’s a medical cardioversion or a direct current cardioversion have a real embolic rate, maybe as high as 1% to 2%. Now that sounds high, but it’s actually relatively low, so to do a prospective trial with a 1% to 2% endpoint, you would have to do tens of thousands of patients.

Now the data that’s been published, that’s the best, actually came from the RE-LY trial with dabigatran. It was published in Circulation12 where they actually had 1200 patients and 1900 cardioversions. The stroke and systemic embolism rate with the 110 mg dose twice a day was 0.8%. With the 150 mg twice a day dose of dabigatran was 0.3% and with the warfarin was 0.6%.  So in spite of this not being a randomized prospective trial, this was at the investigator discretion, there is no signal here that dabigatran has a higher or lower embolic rate than warfarin. In fact, although the hazard ratios are large here and they are not statistically different; the dabigatran had half the stroke rate for warfarin within higher budgets.
Now I would point out that I’ve seen anecdotes and case reports of patients on dabigatran properly treated who have had an embolic event following a cardioversion. In spite of the limitations of this trial, or whether this trial is true or not, I think we’re still going to see some emboli, just like we saw in RE-LY, that there were actually twice the embolic rates in warfarin than on the higher dose of dabigatran.

For rivaroxaban and apixaban, there are some smaller data sets available. Some of this is in abstract form and has not been published, but there are no data in those trials, even though the numbers are much lower, to suggest that there are more embolic rates with these drugs than there are with warfarin. Those trials were a little more limited being double-blind, double-dummy that again, left to the investigator’s discretion less people went on to have, for example, open-label dabigatran cardioversions like we did in RE-LY.
So there’s just not as much data with the other two drugs. You might argue dabigatran has the best data out there, but the data that are least known, and a lot of this will be presented in future meetings, could be similar with the other drugs. No difference, but they are very small data sets.


Dr. Reiffel:Thanks Jerry. Let’s come to the final section and I’m going to ask Jon to lead this off. We’ve talked about the limitations of warfarin. We’ve talked about the pharmacology of the new agents. We’ve talked about the major clinical trials that have led to their approval for two of them and in the approval process for the third. Now we’ve got a patient sitting across from us that we have not yet had on anticoagulant or we have a patient who’s been on warfarin. Jon, how do we decide which agent to use or whether or not to change somebody who is already on warfarin?


Dr. Halperin: The question of whether or not to change gets to this issue of so-called, “starters” and “switchers.” We know from the broad literature that patients during the first year of anticoagulant therapy with warfarin who have not previously been anticoagulated face about three times the bleeding risk of those who have been managed reasonably well with warfarin for some period of time. So there’s a “run-in” phenomenon that occurs during anticoagulation where people with problems sustaining therapy tend to fall-off treatment because of bleeding or events of one kind or another and, after a while, this leaves a select population. When this occurs, in general the event rates tend to be lower.
What is less clear is whether we can utilize that to make clinical decisions for, say, the patient who is doing just fine on warfarin. Does it make sense for them to change to one of the newer oral anticoagulants? Well, certainly we have the matter of convenience and the lack of need for routine coagulation monitoring. I will distinguish that from lack of monitoring at all, because we do need to keep an eye on such things as renal function, and this “run-in” phenomenon makes it very difficult.

I think the strongest argument in favor of making the change, though, is that results thus far reported show uniformly that each of the new oral anticoagulant drugs offers a substantial reduction in the risk of intracerebral hemorrhage compared to warfarin. Even when anticoagulation intensity with warfarin is maintained within the appropriate therapeutic range, there appears to be an increased risk of intracerebral hemorrhage compared to these new agents. That’s the strongest argument, clinically, that I can think of for making the change.
When it comes to choosing among the drugs, then we have to fall back on the limitations that you pointed out, Dr. Reiffel—the fact that these really cannot be subjected to cross-trial comparisons. We do know some things, though. We know that there are three key similarities. First, all of these new drugs proved noninferior to warfarin for preventing total stroke rates and systemic embolism. All reduced the risk of intracerebral hemorrhage. I believe that’s the strongest finding, the most concordant finding across the three trials reported to date. The reductions in mortality are pretty comparable at about 10% per year.



There are some differences that I think we can be conclusive about. Dabigatran in the 150 mg twice daily dose reduced ischemic stroke more than warfarin, but caused more extracranial bleeding, as I mentioned, particularly in the elderly. Rivaroxaban, given once daily, 20 mg for most patients, reduced fatal bleeding, but caused more extracranial bleeding. Apixaban given twice daily in the main dose of 5 mg twice daily, reduced major bleeding and all-cause mortality, but not cardiovascular mortality or rates of ischemic stroke compared to warfarin. So, there are a lot of uncertainties in this field that we have to live with and we can talk about that next.

Dr. Reiffel: Thanks Jon. So for the reader—if you have a patient sitting with you and you’re trying to decide what to use, how closely do they match the population in the clinical trials? What’s the renal function? What other agents are the patients taking? They’re all affected by P-glycoprotein, transport inhibitors and inducers; and, by the way, it is not just prescription drugs that interfere. Some of the herbals—ginkgo, St. John’s wort, and others—can also do so. Thus, one needs to go through the package insert when talking with patients. Also, what’s the patient’s GI history? The biggest limitation with dabigatran is the GI side effects, the gastroesophageal reflux disease–like symptoms. So, as the new guidelines in Stroke say, I think there is a role for each of these new agents and the decision has to be made considering the patient, the patient’s concomitant drugs, and the patient’s medical history all taken into consideration at the same time.


I think with that, we’re just about out of time. So I think the best conclusion is the new guidelines.10  There is a role for each of these agents and the clinician is going to have to become familiar with the pros and cons of each of these, the drug interactions with each of these, and the dosing issues with each of these. And finally, none of these require anticoagulation monitoring, as Jon pointed out, but one can use blood tests to determine whether patients are compliant or whether the drug has been washed out adequately prior to a procedure. The factor Xa inhibitors do affect the INR to a significant degree. So if the INR is normal, the patient has not recently taken a dose of the drug. The direct thrombin inhibitors, in this case dabigatran, affect the activated partial thromboplastin time, the ecarin clotting time, and the thrombin time. If those are normal, the patient has not taken a recent dose of dabigatran or the drug is washed out. So the pharmacology becomes important for all clinicians to become familiar with if they’re going to prescribe these agents. But the fact that we have effective agents that have no significant dietary interactions, that have far fewer drug interactions, that are noninferior, and for two of them, superior to warfarin, I think it’s important to begin to use them in a more widespread fashion than we have. Warfarin still has the bulk of the marketplace in this arena, and I personally think that using it so often is not providing the best care to patients. These are new exciting agents and we should be moving toward them.

So unless there are final comments from the three of you—last chance to weigh in—I would like to say thank you to each of you for participating, and I hope that the readers find this roundtable useful in their care of patients.

Dr. Kowey: Jim, I think this practical approach should be very helpful to clinicians. I think, as I said in the very beginning, the big problem here is that we’re all going to be deluged with information about these new drugs. A lot of trial data, a lot of pharmacology information—so it is very valuable for all of us to drill down to the hard facts that will help us make the right decision for each of our patients as we did today.

Dr. Reiffel: Thank you all for taking time to take part in this discussion.