Dr. Reiffel: I’m Jim Reiffel, MD, cardiologist and electrophysiologist from Columbia University in New York City. With me are Drs Halperin, Kowey, and Naccarelli.
Dr. Halperin: I’m Jonathan Halperin a cardiologist at the Mount Sinai Medical Center in New York.
Dr. Kowey: I’m Peter Kowey, cardiologist at Lankenau, Main Line Health, and Jefferson Medical College in Philadelphia.
Dr. Naccarelli: I am Jerry Naccarelli from the Penn State University College of Medicine at Hershey.
Dr. Reiffel: Today we’re going to be talking about the role of and selection process for the new anticoagulants in nonvalvular atrial fibrillation. To start us off, Jon, maybe you could comment on what do we mean by nonvalvular atrial fibrillation as it pertains to three studies: Randomized Evaluation of Long Tern Anticoagulant Therapy (RE-LY),1 Rivaroxaban Once-Daily Oral Direct Factor Xa Inhinition Compared with Vitamin K Antagonism for Prevention of Stroke and Embolism Trial in Atrial Fibrillation (ROCKET AF),2 and Apixaban for the Prevention of Stroke in Subjects With Atrial Fibrillation (ARISTOTLE).3
Dr. Halperin: These studies were all based on noninferiority designs, which means that they are basically comparing an active treatment—warfarin, the standard of care—to a new agent without a placebo control. Since that requires the extrapolation to efficacy against a putative placebo, such trials require parallelism or a constancy assumption with respect to the trials conducted years ago that compared warfarin to an inactive control, typically a placebo.
The inclusion criteria for the studies, excluded patients with rheumatic mitral stenosis of undefined severity, and any prosthetic heart valve, whether biological or mechanical. It was not stipulated in those days whether, for example, valve repair will have qualified the patient as "valvular atrial fibrillation" and excluded the patients from the trial, because so few of those operations were performed. Also excluded were patients with thyrotoxicosis at the time of atrial fibrillation. So, essentially, patients with rheumatic mitral stenosis, any valve surgery, or thyrotoxicosis, would not fall under the umbrella of nonvalvular atrial fibrillation.
Dr. Reiffel: It is important to emphasize that nonvalvular atrial fibrillation for purposes of selecting a new oral anticoagulant is not the same as not having valve abnormalities. Therefore, if you’ve got a patient with mild or moderate aortic disease or mitral regurgitation—for these purposes, that would not be considered significant valvular heart disease. Patients with valve disease that did not need an operation within the next year were allowed into these trials. Nonvalvular atrial fibrillation does not mean that you can’t have any form of valve disease.
Jerry, Jon mentioned the prior trials with warfarin as the background, so maybe you should talk briefly about the pros and cons of warfarin and why even bother thinking about the new agents?
Dr. Naccarelli: Prior to the release of the newer anticoagulants, the mainstay of therapy for preventing stroke with atrial fibrillation was the use of warfarin: this was based on a series of trials, the largest being Stroke Prevention in Atrial Fibrillation 1 (SPAF1)6 showing that if properly used, warfarin would reduce the risk of stroke by about two-thirds. This requires keeping warfarin in the therapeutic range, which has been found over the years to be 2.0 and 3.0 of an international normalized ratio (INR). Those patients with subtherapeutic INRs obviously have a higher risk of stroke, and patients with supratherapeutic INRs have a higher risk of bleeding.
Now these data have been known for more than 20 years and the issues relate to the following: First, in spite of these pretty impressive data, only about half of the patients in the United States who should be on warfarin by anyone’s guidelines are on warfarin. That of those patients taking warfarin, only about half of the patients in a community-based INR center keep within the therapeutic range. We do a lot better with our anticoagulation clinics where two-thirds of the patients are in the anticoagulation range.
Warfarin is a drug that works. It’s very effective, but the limitations of the use of this drug, have made us eager to have some options. The drug has a slow onset of action. There’s a lot of genetic variation in metabolism. Since warfarin affects the vitamin K dependent pathways of the coagulation cascade, there are significant food and drug interactions. Anything you eat or take will talk to warfarin and have an impact on whether you’re therapeutic or not. The therapeutic windows may narrow and we have to measure something—an INR—in order to try to keep people within a therapeutic range. This requires frequent coagulation monitoring. The dose requirements change depending on how much vitamin K is in the patient’s diet and what medicines they are taking.
Frankly, there are a lot of drug-related patient medical characteristics and genetic issues that we understand, but really aren’t totally incorporated into the use of this drug. It just has become complex. So there’s been a real need of having an alternative anticoagulant that would be easier to use without having to go to an INR clinic or have dietary interactions.
Dr. Reiffel: That’s a good background. Let’s move into the discussion about the new anticoagulants. Peter, can you tell us a little bit about the pharmacology of dabigatran and rivaroxaban and apixaban, the newer anticoagulants—pharmacokinetic/pharmacodynamic interactions and the like?
Dr. Kowey: I think one of the things that is important for physicians to understand is that, as Jerry said, we’ve had a long history of using a very complicated drug, which was warfarin. The new drugs are a significant advance, but they’re not necessarily going to be a cakewalk. They’re all complicated drugs that have their own individual pharmacology including drug interactions and particular pharmacokinetic/pharmacodynamic profiles that physicians are going to have to learn.7
One of the things I think we’ve seen now with the dabigatran experience is that bleeding is much more likely to occur when the drug is used incorrectly. Doctors are going to have to pay careful attention to dosing instructions. This will inevitably be true for the other newer agents as well. But, these drugs are different, as you said, not only in terms of the mechanism of action—dabigatran being a direct thrombin inhibitor and the other two drugs being factor Xa inhibitors—but, also in terms of their pharmacology. For example, each has a different half-life. The half-life of dabigatran is somewhere between 12 to 17 hours. Rivaroxaban is in the 6 to 10 hour range and apixaban is about 8 to 15 hours. Rivaroxaban is the only one of the three drugs that’s dosed once a day. The other two drugs are twice daily in the clinical trials that we’ll talk about.
The other thing that’s different about the drugs is their elimination. The highest amount of kidney elimination is for dabigatran. It’s about 80%, whereas with rivaroxaban, it’s somewhere in the range of about 60% to 65%—70%, and apixaban about 25%. The remainder of the elimination for rivaroxaban and apixaban is through hepatic metabolism. They are both acted upon by the CYP3A4 metabolic pathway. So this obviously leads to issues with regard to drug interactions. For example, concomitant use of ketoconazole, which is a potent 3A4 inhibitor, is problematic.
For dabigatran, we worry more about drugs that have an effect on the P-glycoprotein transport system that can either cause lower levels of dabigatran if those drugs are inducers like rifampin, or may lead to much higher levels in the case of drugs like amiodarone and dronedarone when there is the potential for inhibition of the P-glycoprotein transport system.
Without getting into much more detail, Jim, I think the most important thing for physicians to understand is that there is some burden to digest the pharmacology of these drugs (excuse the pun, by the way) before they’re used in clinical practice. Because if they are not used properly, you get either more clotting or more bleeding.
Dr. Reiffel: I think, Peter, those are all important points. One of the things to comment on, at least as related to the potential complexity of this, using dabigatran as an example, is where you said, the P-glycoprotein inhibitors and inducers, but not the CYP3A4 agents can produce interactions. If you have good renal function, say a clearance of 100 cc a minute, and you drop down to 50 cc a minute, the serum concentrations of dabigatran will go up somewhere between two- and three-fold. In the clinical trials there were no issues in that range in the patients who were enrolled. But if you take the lower end of that range, and then add an agent that can double the serum concentrations, the bleeding risks become substantial. That is why the package insert for dabigatran cautions, for example, against using it concomitantly in the 150 BID dose with an agent such as dronedarone.
Dr. Kowey: Jim, I think it’s very important to get the dose right. We’re a little bit handicapped that we don’t have the 110 dabigatran dose in the United States, but we do have a 75 mg dose. I completely agree with you. I think as you get down into these borderline creatinine clearances, if you do add drugs that potentially may have an effect on elimination, it’s probably a good idea to step down the dose. For rivaroaxaban, we have the 15 mg dose. For dabigatran we have a 75 twice daily dose, which is probably a little bit lower than we’d like it to be, but it is usable in certain clinical circumstances.
Dr. Reiffel: Why don’t we talk a little bit about where those doses came from. They are from the three pivotal trials, RE-LY, ROCKET-AF, and ARISTOTLE. So, Jon, would you tell us a little bit about RE-LY, and then Jerry, ROCKET, and Peter, ARISTOTLE?
Dr. Halperin: RE-LY was an open-label comparison of dabigatran, the direct thrombin inhibitor compared to warfarin in 18,113 patients who had nonvalvular atrial fibrillation and at least one additional stroke risk factor from the CHADS2 risk scoring system.1 Patients were randomly assigned to either open-label warfarin at an INR of 2 to 3, or to one of two blinded doses of dabigatran, 110 mg twice daily—a dose not subsequently approved by the Food and Drug Administration (FDA) for use in the United States—and a higher dose of 150 mg twice daily, which has been approved. There were approximately 6000 patients in each group. The primary objective was noninferiority compared to warfarin for the primary outcome of all stroke—including both ischemic and hemorrhagic stroke—and systemic embolism. The minimum exposure of the patients was one year, the maximum about three years, and the mean exposure about two years. The key safety measure was bleeding during treatment with these drugs.
The primary outcomes were a “home run” for the sponsor of the trial. The higher dose of dabigatran was statistically superior to warfarin for prevention of the primary events—event rates of 1.7% per year with warfarin and 1.1% per year with dabigatran, 150 mg twice daily. And the lower dose, 110 mg twice daily, proved noninferior, but not superior to warfarin with an event rate of 1.5% per year. The criteria for major bleeding events were defined differently across the trials. In the RE-LY trial, this was defined as clinically overt bleeding associated with a hemoglobin drop of 2 grams per deciliter or more, transfusion of 2 or more units of blood, or bleeding at a critical anatomical site. Here, the lower the dose of dabigatran proved safer and the higher dose of dabigatran noninferior to well-adjusted warfarin.