NOAC Update on Dosing and Transitioning

Samuel Z. Goldhaber, MD; Christian T. Ruff, MD, MPH


October 16, 2014

This feature requires the newest version of Flash. You can download it here.

Samuel Z. Goldhaber, MD: This is Dr Sam Goldhaber, for Medscape Cardiology at the European Society of Cardiology (ESC) meetings in Barcelona. I am here with my friend and colleague Dr Christian Ruff, who is also from Brigham and Women's Hospital and Harvard Medical School. Dr Ruff is going to enlighten us on the late-breaking controversies with novel oral anticoagulants (NOACs).

Let's start with the background of what is the biggest controversy right now with the NOACs. They were introduced as anticoagulants that could be given in a fixed dose without any laboratory monitoring whatsoever. Then, a real blockbuster of an article was published by Reilly and colleagues[1] in the Journal of the American College of Cardiology on dabigatran. Could you review the take-home points of that study, and then tell us what you have done with edoxaban in the ENGAGE trial?[2]

Christian T. Ruff, MD: You have hit on a hot-button issue for anticoagulation in atrial fibrillation, because all of the NOACs were developed to avoid routine monitoring, which has been the bane of our existence with warfarin and other vitamin K antagonists for 50 years. None of the NOAC agent studies included routine monitoring in the trial design. All of these agents were given in dosing. Some allowed dose reductions to adjust for factors that can increase bleeding.

The RE-LY team and investigators and Boehringer Ingelheim published a very nice article[1] looking at the exposure to dabigatran and the risks for both stroke and bleeding. They found that dabigatran exposure was highly correlated with both age and renal function. The older you were and the more impaired your renal function, the higher your blood levels of dabigatran. They found that if you looked at concentrations of dabigatran and outcome, as the concentration increased, so did the bleeding, but stroke risk was reduced. Relatively, however, as the concentration increased, there was more bleeding than additional stroke protection.

A conclusion of that study, as well as the press reports that followed, was that if you allow patients to have a high level of drug, that poses an undue bleeding risk. The bleeding risk needs to be modified, or you need to lower the exposure concentration.

One of the questions was whether we should be monitoring these drugs. Should we be measuring drug levels? Should we be measuring anticoagulant activity to us select the dose that will optimize an individual patient's risk for bleeding vs risk for stroke? That is completely opposite to what was done in the trials, which conducted no routine monitoring. We also thought that the NOACs would provide us with a more convenient and safer option for anticoagulation.

As studied, without routine monitoring, all of the NOACs were as effective, if not more effective, than warfarin and reduced the risk for serious bleeding, life-threatening bleeding, and intracranial hemorrhage significantly when used for the indication of stroke prevention in atrial fibrillation.

Dr Goldhaber: Long before the Reilly paper was published, you planned ancillary trials with edoxaban and monitoring drug levels in ENGAGE.

Dr Ruff: We knew in advance from phase 1 and phase 2 work with edoxaban that some very common clinical features predicted both increased edoxaban exposure and increased risk for bleeding. These features included impaired renal clearance; low body weight; and such concurrent medications as P-glycoprotein inhibitors, which we know boost the levels of many of these NOACs. In advance, we mandated in the protocol that all patients who had a creatinine clearance ≤ 50 mL/min or body weight ≤ 60 kg, or were taking a potent P-glycoprotein inhibitors, received one half of the dose of the NOAC.

Dr Goldhaber: What are the most common P-glycoprotein inhibitors?

Dr Ruff: The most common agents used in clinical practice are verapamil; dronedarone; and antibiotics, such as erythromycin. In the ENGAGE trial, the high-dose regimen started at 60 mg, so patients with a 50% dose reduction received 30 mg. In the lower-dose regimen, patients who started out on 30 mg would receive 15 mg. Of interest, more than one third of patients in the trial were on a reduced dose (25% at randomization and nine percent during the trial).

An important question is whether this clinical dose decision worked. Did we prevent the excess of exposure that we anticipated in these patients? The answer presented at the ESC Hot Line Session[2] is "yes." When we looked at the edoxaban concentrations in patients who were dose-reduced and their anti-factor Xa activity, not only did we prevent an excess exposure—we ended up with slightly lower blood levels of edoxaban, and slightly lower levels of anti-factor Xa activity.

But the more important question is about outcomes. We prevented excess exposure in these high-risk patients, but how did they do? Did they have more strokes? Did they have more bleeding? Very reassuringly, when we looked at stroke and systemic embolism, the primary efficacy endpoint of the trial, we found that even though we reduced the dose by one half in both the high- and low-dose regimens, the efficacy was the same. Even with one half the dose and lower concentrations, efficacy was preserved.

This was an interesting and unexpected finding. The study would have been a win if we had the same bleeding rate, because these patients are at high risk for stroke and bleeding, just by having these clinical characteristics. Not only did we see similar relative bleeding, we saw markedly less bleeding. By lowering the concentration, we achieved similar efficacy and less bleeding.

The question is, how does that happen? We modeled the relationship between edoxaban exposure and outcomes and found, for lack of a better word, that the therapeutic window is different for bleeding and stroke. It's flatter for stroke, so as you increase the concentration, stroke rates fall, but you don't get a lot of bang for your buck by pushing the dose. For bleeding, however, when you increase the concentration, you pay a relatively higher price.

Dr Goldhaber: Isn't that what Reilly and colleagues found with dabigatran?

Dr Ruff: That is exactly what they found. This has now been found with dabigatran. It has been found with the factor Xa inhibitors. An important difference between the RE-LY trial and edoxaban is that in the RE-LY trial, there was no dose adjustment.

Many people concluded from the Reilly study that we need to monitor the NOACs anticoagulant activity or measure dabigatran concentration to lower the dose. We found that you don't need to monitor anticoagulant activity or drug concentration. You can simply use clinical features, which you can identify in advance, and reduce the dose in those patients. We tested clinical features alone, with no routine monitoring, and that seemed to work.

Dr Goldhaber: Basically, in advance you figured out who needed the lower dose, and then you validated that with edoxaban levels.

Dr Ruff: Exactly. We knew in advance the features that would predict the vulnerable patients who would be at high risk for bleeding. We dose-adjusted them prospectively. After that, we checked their levels, and we found out that clinical system algorithm worked on its own. We didn't need to monitor the levels.

Dr Goldhaber: This would provide a lot of reassurance, to be able to give a fixed dose and then adjust the dose according to the parameters in the ENGAGE trial without ever doing any laboratory coagulation monitoring, wouldn't it?

Dr Ruff: Yes, and there are two key points in what you said. This provides, in my opinion, data that we do not need routine monitoring of concentration or routine monitoring of anticoagulant activity to use these drugs safely in patients.

The other thing is that logistically, it would be very complicated and challenging to figure out what you would use to measure. We found that even in a clinical trial, setting the concentrations vary widely. Picking a specific threshold or cut-point for clinical use would be very hard. It would be a step backward.

The NOACs are a major advance because you don't need routine monitoring, but that doesn't mean that you can forget about the patients. You should be measuring creatinine clearance. You should be seeing them in clinic. You should be checking their body weight. You just don't need to be measuring their anticoagulant activity or their drug concentration.

Dr Goldhaber: Your group also presented information on the relationship between major bleeding in the setting of edoxaban and stroke prevention and mortality in atrial fibrillation.

Dr Ruff: Right. Whenever you prescribe an anticoagulant, you can cause bleeding. There is a real concern in the clinical community about minimizing bleeding risk. It's interesting that although we gave two potencies of regimens for anticoagulation—a high-dose and low-dose edoxaban—they both reduced cardiovascular mortality significantly. In fact, the low dose reduced all-cause mortality significantly, and there was a trend for the high dose. We began to ask: Why are these patients living longer with edoxaban? What is driving the mortality benefit?

A colleague of mine, Dr Robert Giugliano[3] from the TIMI study group, looked at the excess deaths in the warfarin arm and tried to understand what was driving that. His findings were very interesting. Approximately one half of the deaths were caused by fatal bleeding, with the bleeding directly contributing to death in the causal pathway. The patient with major bleeding ended up in an intensive care unit and succumbed to a complication from that event. That was about one half of the excess major bleeds.

He then said, "Out of all the excess deaths in the warfarin arm, how many major bleeds weren't directly related to death?" Approximately 85% or more of the excess deaths in the warfarin arm were in patients who had major bleeding at some point during the trial. We found that among patients who had major bleeding and didn't die of it, more than 90% stopped therapy at least temporarily, and more than 60% of patients never went back on anticoagulation.

Dr Goldhaber: They might die of a stroke.

Dr Ruff: That is exactly what happened. Those patients were at fivefold or higher risk for such ischemic events as stroke, myocardial infarction, and death. Even if patients don't die of a bleeding event, if it causes them to stop their therapy, that is very dangerous.

Dr Goldhaber: A bleeding event is a very bad prognostic sign. Tell us about how you safely transition warfarin to edoxaban and edoxaban to warfarin.

Dr Ruff: A nice thing about clinical trials is that you learn from each other. We knew from the other factor Xa inhibitor trials that although the trials themselves were terrific and successful, at the end of the trial when patients went from their NOAC to warfarin, there was a threefold excess risk for stroke or bleeding. The lay press began to be concerned about a rebound effect—if you stop these new drugs, maybe you are at an increased risk for stroke.

We developed a transition strategy at the end of the ENGAGE trial, learning from the work from our colleagues' patients who were going from edoxaban to open-label warfarin or another vitamin K antagonist.[4] We had a very aggressive algorithm to get them therapeutic as fast as possible. In our edoxaban study, more than 85% of the patients had a therapeutic international normalized ratio (INR) within 2 weeks and almost 99% of them within a month, which is a real credit to our investigators.

Of note, we also overlapped them with a modified edoxaban kit. We didn't just give them their warfarin and allow them to become therapeutic whenever they could get there; we actually gave them a 2-week overlap to protect them from the increased risk while they were subtherapeutic. We found no excess stroke in any of the patients transitioning to open-label anticoagulation. Unlike other overlapping or bridging studies, we had no excess bleeding risk. This provides reassurance that patients can switch anticoagulation in routine practice owing to adverse events, insurance issues, or concomitant medications, and we proved that you can do this safely.

Dr Goldhaber: This is all very important late-breaking information, Christian. Thank you so much for sharing it with our Medscape audience. This is Dr Sam Goldhaber, signing off for Medscape.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.
Post as: