Experts compare new oral anticoagulants preparing to battle for the warfarin market

December 14, 2010


New York, NY - With the recent initial presentation of the ROCKET AF trial with the factor Xa inhibitor rivaroxaban (Xarelto, Bayer/Johnson & Johnson), it appears that there will probably be two new oral anticoagulants available in the not-too-distant future that can be used as an alternative to warfarin in the prevention of stroke in AF patients.

Warfarin, with all its drawbacks in terms of interactions and monitoring requirements, has remained the "gold-standard" treatment for AF patients for decades, with no other therapy matching its efficacy rates without serious side-effect issues. Now, however, it seems Christmas has come, first with dabigatran (Pradaxa, Boehringer Ingelheim), a competitive thrombin inhibitor, showing excellent results in the RE-LY trial, followed a year later with another positive set of results with rivaroxaban in ROCKET. And there are a host of other similar agents in various stages of development, meaning that a group of easier-to-manage drugs may soon be available to replace warfarin.

But, as is inevitable, the question of which of leading agents may be preferable is now looming large. heartwire asked several leading lights in the field their opinions about this issue and the prospects for other similar agents in development.

The drugs and trials so far

In AF, the first to market in several countries was dabigatran, after the RE-LY trial showed the higher 150-mg twice-daily dose to be superior to warfarin, reducing stroke/peripheral embolic events by 34% and the risk of hemorrhagic stroke by 74%, with comparable major bleeding rates. The lower 110-mg dabigatran dose was noninferior to warfarin in terms of efficacy, with a 20% reduction in major bleeding.

In the ROCKET AF trial, reported at last month's AHA meeting, rivaroxaban 20 mg once daily was shown noninferior to warfarin in reducing all-cause stroke and non-central nervous system (CNS) embolism in AF patients, with a similar rate of major bleeding. However, superiority was not achieved in the primary intention-to-treat analysis, although the secondary "on-treatment" analysis did show superiority.

Apixaban (Bristol-Myers Squibb/Pfizer) has been shown to be better than aspirin alone in the AVERROES trial in AF patients unsuited to warfarin. A large-scale trial against warfarin in AF ( ARISTOTLE ) is now under way.

Other agents in late stage development in AF include edoxaban (Daiichi-Sankyo), TAK- 442 (Takeda), betrixaban (Portola/Merck), and darexaban (Astellas). These agents have also been developed for the treatment/prevention of venous thromboembolism (VTE), and many are in trials for the treatment of acute coronary syndromes.

Dr Deepak Bhatt

Many experts who spoke with heart wire stressed the huge accomplishment of actually being able to replace warfarin after all these years. Summing this up, Dr Deepak Bhatt (Brigham and Women's Hospital, Boston, MA) said: "The success of these new drugs means that warfarin will eventually become redundant. This might take several years, but its days are now definitely numbered, and that is a major breakthrough. That is much more important than which drug is better than the others. That is a much harder question to answer and at the moment is based more on opinion than fact."

While it appears from first impressions that dabigatran has the edge over rivaroxaban as it showed clear superiority over warfarin in RE-LY, several experts say it isn't anywhere near cut and dried. The reasoning has been heard many times before—that different drugs in different trials cannot be compared and that the only way to know for sure would be a head-to-head trial, which is very unlikely to be conducted in the near term because it is not in the interests of the companies involved.

From a clinical point of view, I think most doctors would say rivaroxaban looked better than warfarin.

Bhatt, who is involved in the ACS trials of rivaroxaban and apixaban, says it is difficult to choose one agent over the other from the data available so far. "We know that dabigatran was better than warfarin in RE-LY. The results went way beyond noninferiority, which would have been enough, given its lack of requirement for monitoring, but to have shown superiority was really the icing on the cake. But ROCKET AF was terrific as well. It definitely showed noninferiority, and there was a strong suggestion of superiority. Putting all the statistics aside, from a clinical point of view I think most doctors would say rivaroxaban looked better than warfarin—if it were me being treated I would prefer to be on rivaroxaban rather than warfarin."

Reduction in intracranial hemorrhage a key benefit

Dr Alexander Turpie

Dr Alexander Turpie (McMaster University, Hamilton, ON), who has been involved in some aspects of the development of most of these new drugs, takes a similar stance. "Quite clearly, these drugs are an enormous step forward. The data so far indicate that they are all working. I don't think there is a great deal of difference between dabigatran and rivaroxaban. Both drugs reduced intracranial hemorrhage compared with warfarin. This is very important. With all AF patients you know there is a one in 200 to 300 chance of having an intracranial hemorrhage on warfarin. This is halved with dabigatran, and rivaroxaban showed a similar reduction in ROCKET. Dabigatran also showed superiority over warfarin in the primary end point, so there is no contest here—better efficacy and reduced risks of intracranial hemorrhage: win-win. Rivaroxaban was also definitely effective in ROCKET, but I would say the superiority data on rivaroxaban are perhaps not quite as clear."

Dr Lars Wallentin (Uppsala Clinical Research Center, Sweden), cochair of the RE-LY trial, says: "Dabigatran has definitely got the head start and good data, but there will be opportunities for other drugs in this field. In RE-LY, about 20% of patients discontinued treatment for some reason or another, so alternative agents will be needed."

Dr Lars Wallentin

He notes that while full RE-LY trial results have been published, ROCKET AF has not, so there are not enough data available yet on the ROCKET AF trial to make meaningful comparisons between the two drugs.

"ROCKET AF is so far difficult to evaluate. There has been only one short presentation and a few slides. There is a tremendous difference between that and the dabigatran data, which have been available for 16 months, published in several peer-reviewed papers and undergone FDA review. Every detail of the RE-LY study has been scrutinized."

Dr Peter Berger

He says the differences between the intention-to-treat and the on-treatment analyses in ROCKET AF are "somewhat difficult to understand. . . . There seems to have been a high event rate at the end of the trial when rivaroxaban was stopped and when patients were all transferred back onto open-label warfarin. So that is confusing."

Dr Peter Berger (Geisinger Center for Clinical Studies, Danville, PA) believes the different results of the intention-to-treat and on-treatment analyses in ROCKET AF could have been due to the 25% dropout rate. "While this is not a larger dropout rate than similar complicated, multiyear-long trials, it still reduced the ability of rivaroxaban to outperform warfarin, since so many patients stopped taking study drug and took open-label warfarin. That likely contributed to the different results from the two analyses."

It will boil down to personal choice

Dr Kenneth Mahaffey

Dr Kenneth Mahaffey (Duke University, Durham, NC), who is on the executive committee of ROCKET, says making the decision between dabigatran and rivaroxaban will come down to the personal choice of the individual physician and patient. "Statistics and p values are not the only thing doctors look at. They will be interested in safety, efficacy, and tolerability. We saw a similar adverse-event rate in ROCKET AF for rivaroxaban and warfarin. But in RE-LY there was more dyspepsia and potentially more MIs with dabigatran. Doctors will also want to know the potential efficacy based on risk assessment for individual patients, and the patients in ROCKET AF were moderate to high risk, whereas in RE-LY they were low to moderate risk."

Dr Harvey White

Dr Harvey White (Green Lane Hospital, Auckland, New Zealand) notes that dabigatran will benefit in two ways. "Being first off the block is a tremendous advantage for dabigatran. Also, it has the potential advantage of having two doses, where a physician could tailor therapy with the higher 150-mg dose for a patient with a high risk of stroke and a low risk of bleeding and the lower dose to a patient with a high risk of bleeding (eg, the elderly or frail)."

Although the US FDA has not approved the 110-mg dose, it did approve a 75-mg dose for patients with renal dysfunction. Wallentin noted that the 110-mg dose has been approved in Canada and Brazil, and he hopes it will also be made available in Europe.

Rivaroxaban advantage of once-daily dose

Mahaffey points out that the once-daily dosing schedule for rivaroxaban is a big advantage over dabigatran's twice-daily regimen. White concurs, saying: "The twice-daily dosing means that adherence will be a real challenge long term, especially as patients with AF usually feel well for long periods of time. If a patient misses one dose, it may not be catastrophic but missing two doses may well be. So rivaroxaban will get a [boost] because it's a once-a-day medication, and patients will like that."

Dyspepsia: A bellyache for dabigatran

Although full published results for ROCKET AF may level the playing field, of the two front-runners, dabigatran appears to have more of an issue with adverse events, particularly dyspepsia, caused by its low pH.

Turpie notes: "The downside of dabigatran is that it causes dyspepsia in about 10% of patients, some of whom will discontinue treatment because of it. This is not a big problem, as we will know very soon who these patients are, and they can stop taking the drug. But rivaroxaban might have an advantage here. Patients do not like having a bellyache. They won't take a drug if that is a side effect. Doctors could be swayed by that."

And what about the MI issue?

The observation of a numerically increased rate of MI with dabigatran in RE-LY has been a major talking point. Mahaffey notes that there was also a numerical increase in MI in the phase 2 trial of dabigatran in ACS (REDEEM), although these were small numbers, and newly released data from RE-LY shows that there was no statistical difference in MI rates between groups. "There is some overlap between the AF population and the coronary heart disease population, and this may put some people off dabigatran. Rivaroxaban does not appear to have this issue."

Noting that warfarin protects against MI, the RE-LY investigators have suggested that the small increase in MI in the dabigatran group may reflect a slightly lower level of protection of dabigatran. But White points out that the fact that MI was not increased with rivaroxaban in ROCKET AF (in fact, nonsignificantly 9% lower) or with ximelagatran (the first oral antithrombin developed that was withdrawn because of hepatotoxicity) in the SPORTIF trials challenges that theory.

Wallentin says the actual MI increase is still very low (an excess of around two per 1000 on dabigatran vs warfarin). "But cardiovascular mortality, stroke, and stroke mortality were all reduced in the dabigatran group, so it seems to be of limited clinical relevance." He adds that in AF, there is a very low rate of MI, with only a minority of patients also having coronary heart disease. He does concede, though, that some may worry about treating such patients with dabigatran.

Turpie is also not particularly concerned about the MI data: "The competition will play this up. But vascular mortality was reduced, so I don't think it can be a very serious effect." Bhatt agrees with this stance but adds that the MI issue may be more important in the ACS population.

Open-label vs double-blind designs

Another difference between the RE-LY and ROCKET AF trials is that RE-LY was an open-label study while ROCKET AF had a gold-standard double-blind design.

Bhatt explained that an open trial can introduce bias, but this is not possible to measure. For example, clinicians may be more careful with patients on a new experimental drug. "They might overreact to a small adverse effect and order more tests and find things that wouldn't have been found otherwise. Or maybe those tests would find a small intracranial bleed that would have become a catastrophic one if left another 24 hours, so the bias can go either way. All sorts of personal decisions can be influenced if you know which drug the patient is taking."

Had the results of RE-LY not have been so great, I think the open-label design would have been more of an issue. But it still does slightly detract from the results.

Specifically on the open-label design of RE-LY, Bhatt points out that the hard end points such as mortality and intracranial hemorrhage were going in the right direction for dabigatran in RE-LY and they are difficult to affect by bias. "Had the results of RE-LY not have been so great, I think the open-label design would have been more of an issue. But it still does slightly detract from the results."

Turpie said he didn't think the lack of blinding would make much difference to the outcome but it would fuel marketing wars.

But White believes the blinding issue to be an important difference between the trials. "In a noninferiority trial it is critical that the trial is conducted to the highest possible standard. If patients don't take their drugs or events are not ascertained or possibly biased in their assessment, it makes it likely that there will be no difference between the two treatments, and hence noninferiority will be shown. Therefore, double blinding is one of the necessary requirements for a noninferiority trial, and [the lack of blinding in RE-LY] could have contributed to the nonapproval by the FDA of the lower dabigatran dose." In contrast, White believes the case for rivaroxaban approval is stronger, since the ROCKET AF trial was double-blinded.

INR control

There have also been some discussions about whether the new drugs will still be preferable to warfarin in patients who demonstrate good warfarin INR control. The achievement of therapeutic INRs was 58% in the warfarin arm of ROCKET AF vs 64% in the warfarin arm of RE-LY. The lower rate in ROCKET AF was said to be due to the higher-risk patient population.

Sweden has the best INR control in the world, but even here patients will gain from dabigatran because of lower risk of intracranial hemorrhage.

Wallentin points out that in a RE-LY analysis of dabigatran's effects according to INR control, there was still an advantage for the new agent. "Overall, we see a relationship with INR control as to the effect of warfarin. This is not at all surprising. It is well-known. The poorer the INR control on warfarin, the more events will occur. The better effect of dabigatran was attenuated somewhat by better INR control. Cardiovascular death was superior with dabigatran in the overall trial, but this was no longer superior when compared with just in-range warfarin. But with regard to the primary end point, there is really no statistical interaction. The rate of hemorrhagic stroke is also reduced substantially at all levels of INR control, so dabigatran is a big winner here. There seems to be a specific downside of warfarin when it comes to intracranial hemorrhage that is not there with dabigatran. Every physician starting a patient on prophylactic treatment with dabigatran will be able to tell that patient that we don't really increase the risk of hemorrhagic stroke. That is very important. Sweden has the best INR control in the world, but even here patients will gain from dabigatran because of lower risk of intracranial hemorrhage."

Similar INR analyses of rivaroxaban in ROCKET AF are not yet available.


Of course, price will be another important factor to be considered when selecting which agent to use. Dabigatran has come in at a US "wholesale-acquisition" cost of $6.75 per day for AF. Cost-effectiveness over warfarin has been demonstrated, but Berger points out that use of dabigatran will be affected by the existence of different payers for medication, INR checks, hospitalizations, and rehabilitation-center/nursing-home stays. "Fragmentation of the US healthcare delivery system will undoubtedly hurt the uptake of dabigatran, even though it outperformed warfarin in so many ways in the RE-LY trial," he says. The price of rivaroxaban in AF has not yet been set.

Totality of data, including ACS

Bhatt suggested that another factor that might help doctors to select an agent could be the totality of data across the spectrum of indications. "If the trials in ACS look good, then that might be added security to use that particular drug in the AF population, as there will be some overlap. And rivaroxaban is further along in ACS in the phase 3 ATLAS-2 trial. If this is positive, that might persuade me to use rivaroxaban rather than dabigatran in AF, as we will have more data on rivaroxaban."

But initial signs are that these drugs may not be as successful in ACS as they are in AF. No large-scale trials in this indication have yet been completed, but one of them—the APPRAISE-2 trial with apixaban—was stopped recently because of an unacceptable bleeding risk in the apixaban group.

Bhatt, who is on the steering committee of both APPRAISE-2 and ATLAS-2, explained that it had been anticipated that bleeding would be increased in the ACS trials where the new anticoagulants are being added on top of aspirin and clopidogrel. "Adding in another potent anticlotting agent to two similar drugs already on board is bound to increase bleeding risk. We already know that bleeding is a major problem when warfarin is added to dual antiplatelet therapy, and now we've seen the same thing with apixaban. But the big question is whether this is worth it for a reduction in events. My gut feeling is that it could be, particularly in high-risk subgroups. ACS patients have a higher level of activated platelets and clot burden than AF patients, so maybe triple therapy will be justified, but the problem is that, in general, bleeding risk and ischemic risk tend to track together."

It's too much to give all three together.

Turpie is not enthusiastic about ACS as an indication for these new agents. "On the surface, it is difficult to understand why the ACS studies are being done. It's not surprising that the bleeding risk is increased when they are given on top of aspirin plus clopidogrel. It's too much to give all three together. Apixaban has now demonstrated that. But ATLAS-2 is continuing, and they are talking about risk/benefit ratio, so there must be some evidence of efficacy. It's not beyond the realms of possibility, I suppose." Turpie pointed out that dosing will be crucial in ACS to limit bleeding risk and suggested that perhaps a lower dose of apixaban may have been better.

But Wallentin, who cochaired the APPRAISE-2 trial, explained that they had selected the lowest dose that showed efficacy in the phase 2 APPRAISE-1 study. "The phase 2 trial tested four doses of apixaban on top of clopidogrel and aspirin. As expected, there was a signal of increased bleeding with increased dose. The second-lowest dose was chosen for the phase 3 study—10 mg once daily. This dose showed a better signal of efficacy, whereas the lower dose (2.5 mg twice daily) showed less sign of efficacy. The trials had the same [data safety and monitoring board], and they were well aware of the risks involved, and it was decided to give apixaban as 5 mg twice daily to avoid the higher peak of once-daily dosing. This was a cautious approach, and the potential risk/benefit ratio was very much under discussion before the trial started. We were expecting a higher bleeding rate, but we were also hoping that this would be outweighed by a reduction in ischemic events. But unfortunately this was not the case."

Wallentin also believes that these agents have a tough challenge in ACS. "There are three phase 2 trials now (APPRAISE-1, ATLAS ACS-TIMI 46 [ATLAS-1], and REDEEM with dabigatran), all consistently showing an increased bleeding risk when these drugs are added on top of aspirin and clopidogrel. This is hard to get away from. And the ischemic benefits are limited, especially in patients getting PCI. It appears that in this population, it will be difficult to add on any further anticoagulation."

Unlike rivaroxaban and apixaban, dabigatran has not yet started phase 3 trials in ACS. Wallentin explained that after the phase 2 REDEEM study showed the same increased bleeding risk by dose for dabigatran as in other trials and only limited signals of benefit, the company put the ACS indication on hold until the results of APPRAISE-2 and ATLAS-2 were available. "I doubt that their enthusiasm has become larger now," he added.

Data in VTE

These agents have mostly been developed first for the prevention and treatment of VTE, different world markets for one or both. And apixaban is awaiting approval for deep vein thrombosis (DVT) prevention postorthopedic surgery.

Turpie summarized the DVT trials for heart wire : in DVT prevention postorthopedic surgery, both rivaroxaban in RECORD and apixaban in ADVANCE have been shown superior to enoxaparin, the current standard of care. Viewing the trials side by side, he said, the efficacy results with rivaroxaban seemed slightly better, but the bleeding slightly worse than with apixaban, but this can be attributed to the timing of administration. Turpie notes: "Rivaroxaban was given six to eight hours after surgery, whereas apixaban was given 18 hours after surgery, and we know that the closer to surgery anticlotting drugs are given, the better the outcome but the higher the bleeding risk. The timing of these drugs is key in orthopedic surgery."

Dabigatran has not shown such good results in the orthopedic setting, being shown noninferior to current standard of care in two studies but failing to meet noninferiority in other trials, Turpie explained. In Canada, both rivaroxaban and dabigatran are approved for this indication, but only rivaroxaban has gotten reimbursement. Because of this, Turpie says rivaroxaban has "taken off" in Canada.

VTE prophylaxis in medical patients could be the largest market, even larger than AF.

He notes that recruitment has also been completed in a large study ( MAGELLAN ) to evaluate rivaroxaban for the prophylaxis of VTE in medical patients (those who are bedridden). "This could be the largest market, even larger than the AF market. There are an awful lot more medical patients in the world than AF patients," Turpie commented.

In the treatment of DVT, rivaroxaban in EINSTEIN and dabigatran in RECOVER have both been shown to be similar (but more convenient) to the standard of care (low-molecular-weight heparin/warfarin), and apixaban and edoxaban are in phase 3 trials ( AMPLIFY and Hokusai , respectively).

Bhatt notes that the venous trials are generally not as large as the AF/ACS studies. "The bar is a little lower on the venous side. But I believe the totality of data will be important. If one agent has better data over the whole spectrum of indications—DVT/AF/ACS—then that may influence people." Turpie adds that while he doesn't think doctors will look at the data outside their own field when choosing a drug, formularies probably will look across the spectrum.

Others in development: Any advantages?

Of the host of other similar agents in development, most look broadly similar to the lead products. Turpie, however, notes that most of these new agents are primarily excreted by the kidney, and plasma concentration rises sharply in renal-dysfunction patients. One agent in development, betrixaban, has an advantage in this regard, as it is metabolized in the liver rather than excreted by the kidney. "This one therefore should be safer in renal insufficiency. That could be a big issue in AF," Turpie says. Wallentin notes that although the FDA has approved a 75-mg dose of dabigatran for use in patients with renal dysfunction, which he says is "a logical step," if there were alternatives less dependent on renal excretion they would probably be more straightforward and easier to use.

Right now there is no clear way to say one is better than the other.

Mahaffey points out that there will be subtle pharmacokinetic and pharmacodynamic differences between them. "There are lots of different statins and ACE inhibitors. There will probably be a place for many of these agents too."

So although dabigatran has set the pace for these agents in cardiovascular indications, as Bhatt says: "All sorts of things are still up in the air." He concludes: "It's still wide open. We need to see the full published ROCKET AF data set, and there will be more data from several other trials in both AF and ACS before too long. These trials will be key in influencing whether doctors lean toward or against one particular agent. But right now there is no clear way to say one is better than the other."

Bhatt is on executive committee of ATLAS-2 (rivaroxaban) and steering committee of APPRAISE-2 (apixaban), and he receives research funding from AstraZeneca, Bristol-Myers Squibb, Eisai, Sanofi-Aventis, and the Medicines Company. Turpie is a consultant to Bayer/ Johnson & Johnson for rivaroxaban and on the steering committee of ROCKET. He is on the speaker's bureau for dabigatran, an investigator for apixaban, and a consultant for Astellis (YM-150), Takeda (TAK 047, and Portola (betrixaban). Wallentin was cochair of RE-LY and chair of REDEEM (both dabigatran), and he is cochair of ARISTOTLE and APPRAISE I and 2 (all with apixaban). Mahaffey is on the executive committee of ROCKET AF(rivaroxaban). He was involved with REDEEM (dabigatran), not RE-LY, and is a consultant for Johnson &Johnson, Bayer, Boehringer Ingelheim, Merck, and Bristol-Myers Squibb. White is on the steering committees for APPRAISE (apixaban) and ROCKET AF (rivaroxaban). Berger is on an advisory board for dabigatran and is an investigator in trials of rivaroxaban and edoxaban.


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.