New agents for AF and stroke prevention: How to position the best drug for the right patient

The recent approval of new anticoagulant agents to prevent stroke in patients with nonvalvular atrial fibrillation is of great interest to the cardiology community.

Newly approved drugs include dabigatran (Pradaxa, Boehringer Ingelheim), a direct thrombin inhibitor; rivaroxaban (Xarelto, Janssen Pharmaceuticals), a factor Xa inhibitor; and apixaban (Eliquis, Bristol-Myers Squibb/Pfizer), another factor Xa inhibitor. Also on the horizon is edoxaban (Lixiana, Daiichi Sankyo), a factor Xa inhibitor that is being studied in the ENGAGE AF–TIMI 48 trial.

The new agents have been welcomed by cardiologists due to well-known issues with warfarin (Coumadin, Bristol-Myers Squibb), the reference standard for oral anticoagulation, such as very long onset and offset, dosing, cumbersome monitoring, and interactions with food and other medications.

This expert round table focuses on these new therapies and how to position which drugs for which patient. Carl J. Pepine, MD, Chief Medical Editor of Cardiology Today, lead the discussion.

Round Table Participants

Moderator

Carl J. Pepine, MD

Cardiology Today Chief Medical Editor

Division Director Emeritus, Eminent Scholar Emeritus and Professor of Medicine, Division of Cardiovascular Medicine, University of Florida College of Medicine

Richard C. Becker, MD, MEd

Director, Cardiovascular Thrombosis Center

Member of the Divisions of Cardiology and Hematology, Duke University Medical Center

Christopher B. Granger, MD

Professor of Medicine, Duke University Medical Center

Director, Cardiac Care Unit

Clinical Trialist, Duke Clinical Research Institute

Albert L. Waldo, MD

Pritchard Professor of Cardiology and Professor of Medicine, Case Western Reserve University

Associate Chief of Cardiovascular Medicine for Academic Affairs, University Hospitals Case Medical Center

Member of the Arrhythmia Disorders section, Cardiology Today Editorial Board

Similarities of the novel therapies

Pepine: A dilemma that has emerged over the last 2 years is how to best position these new drugs relative to patient characteristics. Let’s begin with the question of how similar are these newer agents?

Christopher B. Granger, MD: In terms of outcomes, they are more similar than dissimilar. But, in terms of the drugs, they are fundamentally different. Specifically, dabigatran, a direct thrombin inhibitor, works on a different part of the coagulation system than the factor Xa inhibitors. In fact, it is fairly striking that the outcomes are improved so similarly with both.

The findings from the RE-LY trial were remarkable for these new agents compared with warfarin. Dabigatran, especially the 150-mg twice-daily dose, was significantly better than warfarin at preventing stroke, including ischemic stroke; had about the same rate of bleeding; and had a remarkable more than 50% reduction in intracranial hemorrhage and hemorrhagic stroke. Then, with rivaroxaban we saw similar findings. There was not as significant reduction in stroke, but numerically fewer strokes with rivaroxaban, and a similar reduction, maybe a magnitude less, in intracranial hemorrhage. With apixaban, there were similar findings of a significant reduction in total stroke driven by less hemorrhagic stroke and, again, about a 50% reduction in intracranial hemorrhage. Interestingly, and I believe this relates more to the dose of apixaban studied than a fundamental difference in the drug, there was also a 31% reduction in major bleeding compared with warfarin.

Albert L. Waldo, MD: Of note, the drop-out rate in these trials is remarkably high — over 20% in all the trials. These trials had huge numbers of patients, but original warfarin trials in the ’90s had far fewer patients. For instance, the NIH SPAF I trial had only about 300 patients, whereas ROCKET AF had about 14,000, and trials of apixaban and dabigatran had about 18,000. Had there been 18,000 patients in the ROCKET AF trial, I believe rivaroxaban would also have been superior to warfarin on an intention-to-treat basis. As it is, they just missed it (P=.117).

With huge numbers of patients, small differences become significant.

Richard C. Becker, MD, MEd: As far as the drugs’ performance in clinical trials, there are more similarities than differences. But when one looks at the drugs and the trials, there are a slew of differences that I believe, when all is said and done, will be very important. Let me elaborate.

First, the patient populations in the trials were different. The comment has been made about the ROCKET AF trial and the number of patients, but keep in mind that the average CHADS2 score was 3.5, and 55% of the patients had a prior stroke or transient ischemic attack. These are not easy patients to find, but they are the patients that we typically see.

Although the target, in the case of the direct factor Xa inhibitors, is identical, the pharmacology — specifically, how they are cleared, rate of clearance, dosing frequency and reversibility — is different. In addition, the target binding characteristics, protein binding in the blood and, in turn, the proportion of “free drug” differs among the novel oral anticoagulants.

As a practicing clinician, the message I would like to make clear to other clinicians is that we must familiarize ourselves with these drugs, how they work, their pharmacology and potential drug-drug interactions in order to make optimal treatment decisions. Fundamentally, the goal is the right drug, at the right dose, for the right patient.

Waldo: There are significant differences in renal clearance as well. Dabigatran is 80% cleared by the kidneys. As people get older their kidney function doesn’t get better, so this becomes a real issue. But they all have clearance of some consequence.

Granger: That’s a real strength of rivaroxaban in the ROCKET AF trial. There was a predefined 25% reduction in dose from 20 mg to 15 mg once-daily rivaroxaban for patients with estimated creatinine clearance <50 mL/min, and the results in this group looked very consistent with the overall results. So it looks like it was the right decision. With apixaban, there was also a dose reduction for patients who had a combination of high risk for bleeding and high drug exposure that included two of the three criteria: age ≥80 years; serum creatinine ≥1.5 mg/dL; body weight ≤60 kg. Patients who had two of those criteria had half the dose of apixaban, 2.5 mg twice-daily. And, likewise, that seemed to be the right decision. Looking at that subgroup, it was small numbers, but, if anything, the point estimates were even more favorable for a lower rate of bleeding and a lower rate of stroke with apixaban than warfarin.

I completely agree with Dr. Waldo that kidney function is critically important. Especially for patients with borderline renal function, one needs to be much more careful with these drugs, and probably even more so with dabigatran than with warfarin.

Safety and dosing

Pepine: Given that we have several wonderful new options, I believe that you all shook your heads indicating that they are probably safer from the major risk that we all fear as a practicing cardiologist — creating an intracerebral bleed. Relative to safety vs. warfarin, would you agree that the cardiologist should feel comfortable telling their patients to take one of these new drugs because from the central nervous system standpoint it is probably safer than warfarin?

Becker: I think it is critical to acknowledge that the intracranial hemorrhage signal was present whether the person was older than 75 years or younger than 75 years because we all have patients that span these broad age categories.

Pepine: What about other differences such as frequency of dosing and side effects that the practitioner ought to bring into this discussion with their patients?

Granger: Age is another important factor. In each of these major trials we did approximately 20 or 30 prespecified subgroup analyses. Usually, subgroup analyses are either unhelpful or misleading; of all the subgroups in all of these trials, there is only one that I think is very important. That was in RE-LY, and it was age. As people got older the relative risk for bleeding with dabigatran vs. warfarin became less favorable for dabigatran. For low-dose dabigatran, which has less bleeding than warfarin, the degree of less bleeding shrank as people got older, and for the higher dose there tended to be a bleeding advantage with dabigatran vs. warfarin in younger patients and similar or slightly more bleeding with dabigatran in the elderly. Many of us believe now that we are unfortunately hampered in the United States because we do not have the 110-mg dose of dabigatran. I think for most patients over the age of 80 years, if one is using dabigatran the ideal dose would be the lower dose.

Waldo: Edoxaban also takes into effect the lower dosing. What’s unique about edoxaban is that it has two different doses.

Granger: And edoxaban is being developed with a strategy of changing the dose if patients have changing renal function over time.

Pepine: Would anyone like to comment on the relative merits or demerits of once-a-day or twice-a-day dosing?

Waldo: The issue of dosing is huge. I’ve seen from patients who are taking a drug twice a day that it is not so easy. Three times a day is worse. I don’t get to start too many patients on these drugs in the practice that I have, but I see a lot of patients who have already been on these drugs. If they forget their dose it is a problem because 3 or 4 hours later they remember and know they have to take it, but when do they take their next dose? They are worried that if they take that next dose on time, the blood level will be too high so that they may have a bleed.

Granger: I would say it’s a modest issue, but it creates an important issue for some patients. It’s a real advantage to have a once-a-day option. My mother, for example, is on rivaroxaban and she really likes the once-a-day feature, and I think many patients feel the same way. But, most of this population is taking some drugs in the morning and in the evening. The addition of tools to improve adherence is critically important with these new drugs because many of us are concerned about adherence compared with warfarin, with which we have ways to monitor adherence on a regular basis.

Becker: I’ll comment briefly on once-daily vs. twice-daily dosing and echo the comments made by my colleagues. Patient selection is critical. Some patients will voice a preference and admit that twice-daily dosing may be difficult. The clinician must engage his or her patients in an open discussion to avoid preventable events and complications.

Reversability

Pepine: Richard brought up the issue of reversibility. We know for sure that dabigatran and rivaroxaban may have some reversibility with prothrombin concentrate. Do you want to expand on that, and who do you pay special consideration to for the need for rapid reversibility?

Becker: It is an interesting evolution. It’s the same conversation that clinicians have been having for many years as new anticoagulants have been developed. It started with low–molecular-weight heparins, and then the intravenous direct thrombin inhibitors.

If there’s a complication, how will we address this? Short-acting drugs have an advantage in that regard as compared with longer-acting drugs, but there are certain times where even a short-acting drug may seem like an eternity when somebody has a life-threatening hemorrhagic complication.

For individuals with a history of bleeding events, in some instances a person’s history would suggest that they are not a candidate for an anticoagulant. In other cases, some might say that the net clinical benefit or the risk–benefit still favors treatment. But if a person has had a life-threatening hemorrhagic complication previously or has a comorbid illness that increases risk for bleeding, the potential for reversibility should be included in the equation as to what is the right drug, or dose of that drug, for this particular patient.

I don’t think that it should be the ultimate decision, or the deciding factor, in many patients to say, “We don’t have an antidote for this.” We don’t have antidotes really for any of our anticoagulants when you look at it carefully, with the exception of perhaps protamine for unfractionated heparin. But as we think about the patients carefully, I believe we can decide and make an informed decision about the right drug for the right patient.

Granger: I disagree somewhat with my friend, Rick. A manuscript in Circulation by Healey and colleagues from RE-LY included a report of how patients fared who needed to have either emergency or elective procedures during the course of the trial. First, it’s very common. Twenty-five percent over a period of less than 2 years had a procedure, so this is not a rare phenomenon. This report showed that patients on dabigatran, compared to warfarin, had a shorter duration of time when it was an elective procedure between when they stopped the drug and when they had the procedure. This is not surprising, and they generally followed the recommendations in the package insert of varying that time depending on renal function and risk for bleeding at the procedure. The package insert states: “If possible, discontinue PRADAXA 1 to 2 days (CrCl ≥50 mL/min) or 3 to 5 days (CrCl <50 mL/min) before invasive or surgical procedures because of the increased risk of bleeding.” In that context, the bleeding rate was numerically lower in the dabigatran group than the warfarin group. This was true for emergency procedures and for procedures performed within 24 hours of the last dose of the drug.

This tells us that in spite of lack of reversibility there was no greater and, if anything, a tendency for lower risk for bleeding with the novel agents compared with warfarin. Therefore, some of us believe that this type of clinical outcome data argue that lack of a reversal agent should not be a reason to avoid the use of the new agents because, as Rick points out with warfarin, although we have vitamin K and fresh frozen plasma and prothrombin complex concentrate, those are actually not commonly used and it is also unclear that they make a big difference in improving outcome. For example, with intracranial hemorrhage, which is our most feared complication, and in spite of those reversal strategies, the outcome is poor with intracranial hemorrhage. The best way to deal with intracranial hemorrhage is to prevent it from happening.

For a larger image, click here.

Dangers of serious bleeds

Pepine: While on the topic of intracranial hemorrhage, let’s talk about a couple of scenarios. I believe that there is general agreement across the round table that these drugs are safer than warfarin, however you want to measure that, and each of you commented on the lower rate of intracranial hemorrhage. Given the patient with AF with a history of one spontaneous intracranial bleed, would you consider offering the patient anticoagulation with one of the newer drugs?

Waldo: In my opinion, there are other options. With minimally invasive surgery, a clip (AtriCure) can be placed on the atrial appendage and the patient won’t need anticoagulation. Then there are Amplatzer PFO Occluder (St. Jude Medical) and the Watchman Left Atrial Appendage Closure Device (Boston Scientific). With these devices you still have some need for anticoagulation. For example, for the Watchman device, you need 45 days of warfarin. So, these patients still have to be on warfarin for a while.

Pepine: Would you offer them one of these newer drugs vs. a Watchman or vs. a clip?

Waldo: I would offer them a clip.

Granger: Carl, it’s important to note that not all intracranial hemorrhage is the same, and frequently it might be worthwhile to consult a neurologist if it’s someone who has an aneurysm that was clipped, or it could be somebody who had a severe hypertensive episode and now they’re controlled. There are patients who you would prohibitively deem high risk for recurrent hemorrhage on oral anticoagulation.

Waldo: There are no absolutes here. You have to really know your patient.

Pepine: Let’s take the other patient who has history of an intracranial bleed on warfarin with only minimal residual defect, at the present still in AF, and has been told Coumadin is contraindicated. Would you consider such a patient a candidate for one of the newer drugs?

Waldo: I wouldn’t. I would not first recommend any anticoagulant if I had an alternative. And now there are alternatives.

Granger: I would consider one of the novel oral anticoagulants in selected patients. I would even consider resumption of warfarin for a patient who had an intracranial hemorrhage where one felt as though the risk may not be terribly high of recurrence. But in general any oral anticoagulant would be avoided.

Becker: While intracranial hemorrhage does not occur frequently, the scenario you propose is not rare. One must acknowledge from the outset that patients with a history of intracranial bleeding were excluded from participation in RE-LY, ROCKET AF and ARISTOTLE.

Patients who have experienced a prior intraparenchymal hemorrhage while on warfarin are at particularly high risk with an ‘anticoagulant re-challenge.’ The literature has shown repeatedly that an elevated INR, unlike extracranial bleeding, does not accurately portend risk for a first or subsequent events.

Waldo: It is very important to emphasize that. Also, I wish to point out that, for instance, depending on the etiology of a subdural hemorrhage, it may be quite reasonable to resume oral anticoagulation after an appropriate waiting period.

Classification scores

Pepine: The next issue is CHADS2 score. We are all familiar with the limitations of whatever classification system you choose, but with the emergence of the CHA2DS2-VASc score, what’s your belief in terms of better estimating risk, particularly among women and patients older than 65 years?

Waldo: The advantage of the CHA2DS2-VASc score seems to be that it sorts out at the low end who really is at risk and who really isn’t. If you look back at the 2001 guidelines, they incorporated the things that are in CHA2DS2-VASc. If you were a woman, you were considered high risk in 2001; the guidelines also considered myocardial infarction, not peripheral vascular disease. Age 65 to 74 years was considered a moderate risk. But in 2006 it changed. What CHA2DS2-VASc did was to again include these latter three risk factors, and it gave the age 75 years or older age category 2 points.

The bottom line is that a CHADS2 score of 0 doesn’t discriminate who really doesn’t need anticoagulation and who does. I believe CHA2DS2-VASc does a better job. If you have a CHA2DS2-VASc score of 0 your risk for stroke is really low. In fact, now the European Society of Cardiology guidelines recommend no antithrombotic therapy at all if you have a CHA2DS2-VASc score of 0. If you have a CHA2DS2-VASc of 1, except for women aged younger than 65 years, they recommend an oral anticoagulant. There is no recommendation for the use of aspirin alone. For a CHA2DS2-VASc of 1, if the patient will not or cannot take an oral anticoagulant, then I would suggest a combination of aspirin and clopidogrel.

Granger: Female sex alone, in other words a woman under age 65 [years], is not a reason because that woman seems to be at the same risk as somebody whose CHA2DS2-VASc score is 0. The whole issue of whether to use CHA2DS2-VASc is a complicated one because we can’t even use CHADS2 score very well. Neither is great at discriminating risk. A lot of us are reluctant to think that we should be having a score that has eight factors until perhaps we can have better clinical decisions supporting electronic medical record systems, then maybe it will be helpful.

There are also some fascinating data emerging about biomarkers and risk in AF. For example, either BNP or high-sensitivity troponin seems to be equally effective as CHADS2 score or CHA2DS2-VASc for discriminating risk for stroke for patients with AF. This is not ready to be used in practice, but in the future biomarkers may be helpful to refine prediction of risk.

My third point is the novel agents have really changed risk stratification because at a substantially lower risk for intracranial hemorrhage, and with some of the drugs and doses a lower risk of bleeding, one might argue that the benefits are warranted in a much broader population assuming that we can afford the drugs.

Pepine: So, we don’t really need a score. We can tell the patients who are at very high risk by our clinical evaluation.

Where I have trouble as a practitioner is women who have had hypertension, I feel comfortable with CHA2DS2-VASc because it puts all those women into 2 — the women who are older than 65 years with hypertension — and I feel I have more reason now to approach them with a newer drug that is a little safer. Is this just wishful thinking on my part or do you think there’s something to this?

Becker: I think there is something to it. When all is said and done, you still have to sit down with your patient and have a conversation about their risk for having a stroke. As a clinician, I agree with you. I think that CHA2DS2-VASc is helpful to have that conversation and for a clinician to say, “I believe that I’m doing the right thing.” It doesn’t answer the question about the pharmacology. We can talk about bleeding risk scores and where that fits in, but there has to be a starting point.

Granger: I agree completely. You brought up an important point, which is it’s not a binary decision. You have a patient and you may think that for this patient the benefit outweighs the risk, but then talk to the patient and he or she says, “I don’t want to be on an anticoagulant.” Using something like further discrimination, let’s say the 65-year-old woman with hypertension, then I think that it’s very reasonable for that patient to then be more clear and strong in the recommendation to be on one of the new oral anticoagulants with lower bleeding risk, and to encourage that person to follow with that plan.

Waldo: We’re all thinking the same way. With the CHA2DS2-VASc score, if it is 0 you don’t do anything. With CHA2DS2-VASc, anyone with a score of 1 or more will get an oral anticoagulant. They’ve abandoned aspirin quite appropriately, and so largely have the Canadian Cardiovascular Society and the American College of Chest Physicians. So, maybe the guidelines are really simplified, if you think about it in those terms.

Then there is the issue of the risk vs. net clinical benefit. For the most part, it’s a fivefold risk for stroke compared with bleeding, and most bleeding is manageable. If you think about it in those terms, I think that’s reasonable. We get in the car and risk our lives every day. But we do it because we understand what the risk is, but the benefit far outweighs the risk.

Granger: Just one other point. The AVERROES clinical trial does help reinforce these concepts. Compared with aspirin, and many of these patients were CHADS2 1, there was a consistent 54% relative risk reduction in stroke with apixaban vs. aspirin across the CHADS2 scores in this generally lower-risk cohort. It really reinforces the idea that these oral anticoagulants do provide meaningful benefit in the medium-low risk group with CHADS2 score of 1.

Part two of this round table will appear online in February. 

Expect more expert discussion on when and how to bridge therapy; potential MI signals; cost; and future use of these new therapies.

Richard C. Becker, MD, MEd, can be reached at Duke University Medical Center, DUMC 3850, Durham, NC 27710; email: richard.becker@duke.edu.
Christoper B. Granger, MD, can be reached at Duke University Medical Center, DUMC 3850, Durham, NC 27710; email: christopher.granger@duke.edu.
Carl J. Pepine, MD, can be reached at the Cardiology Today office, 6900 Grove Road, Thorofare, NJ 08086; email: carl.pepine@medicine.ufl.edu.
Albert L. Waldo, MD, can be reached at University Hospitals Case Medical Center, 11100 Euclid Ave., Cleveland, OH 44106; email: albert.waldo@case.edu.

Disclosures: Becker reports research support from Johnson & Johnson and scientific advisories for Bayer, Bristol-Myers Squibb, Boehringer Ingelheim and Daiichi Sankyo. Granger reports research grants/contracts and consulting/other services from Boehringer Ingelheim, Bristol-Myers Squibb and Pfizer. Pepine reports no relevant financial disclosures. Waldo reports consulting for AtriCure, Daiichi Sankyo and Janssen Pharmaceuticals, and is a speaker for Janssen Pharmaceuticals.