Low-Dose Rivaroxaban Looks Good in ACS: ATLAS ACS 2

November 13, 2011

Updated November 13, 2011 (Orlando, Florida) - The lower of the two doses of the new oral anticoagulant rivaroxaban (Xarelto, Bayer/Johnson & Johnson) tested in the ATLAS ACS 2 TIMI 51 trial has shown promising results, with a reduction in overall and cardiovascular mortality vs placebo, despite an increased risk of bleeding and intracranial hemorrhage (ICH) [1].

The trial, presented at the American Heart Association (AHA) 2011 Scientific Sessions today, compared two doses of rivaroxaban with placebo in ACS patients. All patients were taking low-dose (75-100 mg) aspirin and 93% were also on clopidogrel (Plavix, Bristol-Myers Squibb/Sanofi-Aventis). Study treatment was started an average of 4.6 days after the ACS event. Patients with a previous stroke or transient ischemic attack (TIA) were excluded, as this group has been shown to have a particularly high risk of ICH in previous trials of other antithrombotic agents. The population was high risk, with half having had a STEMI.

Both rivaroxaban doses reduced the primary end point of cardiovascular death/MI/stroke, at the cost of increased bleeding rates. The 2.5-mg twice-daily dose had the better benefit/risk balance, due to a lower bleeding risk than the 5-mg twice-daily dose.

"Thus, the addition of very low-dose anticoagulation with rivaroxaban may represent a new treatment strategy in patients with a recent acute coronary syndrome," the ATLAS investigators conclude in the New England Journal of Medicine paper, published online today to coincide with the AHA presentation.

The trial randomized 15 526 ACS patients to one of the two doses of rivaroxaban or placebo for a mean of 13 months and up to 31 months.

The overall results showed a significant 16% reduction in the primary efficacy end point for the two doses combined vs placebo, with a threefold increase in major bleeding and intracranial hemorrhage but no significant increase in fatal bleeding.

Results of Combined Rivaroxaban Doses vs Placebo

End point Rivaroxaban combined doses (%) Placebo (%) p
CV death/MI/stroke* 8.9 10.7 0.008
Major non-CABG bleeding 2.1 0.6 <0.001
ICH 0.6 0.2 0.009
Fatal bleeding 0.3 0.2 0.66

*Primary end point

In terms of the individual doses, the lower dose showed the best results, with a reduction in cardiovascular death and all-cause death as well as in the overall primary end point.

Results of Rivaroxaban 2.5 mg Twice Daily Dose vs Placebo

End point Rivaroxaban 2.5 mg twice daily (%) Placebo


HR (95% CI) p
CV death/MI/stroke* 9.1 10.7 0.84 (0.72–0.97 ) 0.02
CV death 2.7 4.1 0. 66 (0.51–0.86) 0.002
All-cause death 2.9 4.5 0.68 (0.53–0. 87) 0.002
Major non-CABG bleeding 1.8 0.6 3.46 (2.08–5.77) 0.001
ICH 0.4 0.2 2.83 (1.02–7.86) 0.04
Fatal bleeding 0.1 0.2 0.67 (0.24–1.89) 0.45

*Primary end point

ATLAS investigator Dr Michael Gibson (Beth Israel Deaconess Hospital, Boston, MA) commented to heartwire : "We have shown a 32% relative reduction in all-cause mortality with the 2.5-mg dose in the whole population and a 36% reduction in those on dual antiplatelet therapy. This translates into one death prevented for every 56 patients treated for two years. This is big news. Okay, so there is still an increase in bleeding, but the mortality benefit outweighs the bleeding concerns. There was no increase in fatal bleeding, and although there was an increase in intracranial hemorrhage (ICH), this tended to be in the not-so-serious cases. Of the patents who had an ICH, 80% of those on placebo died compared with 30% of those on rivaroxaban. Also, in patients who had strokes during the trial, those on rivaroxaban had less disability than those on placebo," he added

Caution on Patients With Higher Bleeding Risk

In an accompanying editorial in the journal [2], Drs Matthew Roe and Magnus Ohman (Duke University, Durham, NC) caution that the proportion of patients who were elderly or had reduced renal function in ATLAS 2 was small, and these patients had numerically higher bleeding rates. They suggest, therefore, that the current results are an important development for relatively young and healthy patients with an ACS, but they might not be applicable to patients at higher bleeding risks who are commonly treated in routine practice.

But Gibson disagrees with this point. He argued: "We did include elderly and renal-impaired patients. We didn't have a lot of them, but we had enough. And these subgroups also showed benefit from rivaroxaban." He said the trial was not powered to look at mortality in subgroups, adding, "When we see an overall reduction in mortality as large as we did, with no major differences between subgroups, I think it is hard to make a case for not treating patients."

Mechanism of Mortality Reduction--Sudden Death?

Designated discussant of the ATLAS-2 trial, Dr Paul Armstrong (University of Alberta, Edmonton), told heartwire , "There is no question that this is a significant advance. It is the first time that an antithrombotic has shown a reduction in mortality when added to dual antiplatelet therapy. But to what extent that mortality reduction is precise and how it occurs is a bit of a mystery."

He noted that stroke was not improved and there was no reduction in MI with the low dose, "so what is the mechanism of the mortality benefit?" He suggested that it could have been brought about by a reduction in sudden death, and there was a small reduction in stent thrombosis.

Gibson agreed that the mortality reduction probably was brought about by a reduction in sudden death. He noted that there were 43 unwitnessed sudden deaths in the 2.5-mg rivaroxaban group vs 77 in the placebo group. "And we know from the literature that about 70% of sudden deaths are thrombotic, so that is what it must be."

On addressing the question of why the 5-mg group didn't show a mortality benefit, Gibson pointed out that there was a significant increase in fatal bleeding in the 5-mg group compared with the 2.5-mg group. "So we probably lost some of the benefit because of that. In addition, patients in the 5-mg group more frequently had a bleed prior to an event, so some of the deaths not officially attributed to bleeding may actually have been caused by bleeding." He added that patients who bleed generally stop taking their drugs, which leads to an increased risk of an event.

Armstrong also voiced concern about the ICH rate. "This is a worry, and survivors of ICH often have significant residual impairment. I need to know more about these events before making up my mind." He added that doctors would need guidelines as to which patients to give rivaroxaban, and these would likely be focused on baseline bleeding risk. But in general he was positive. "I wouldn't use this drug in patients with a previous stroke/TIA or gastric bleed. But I would consider it in most others," he said.

Dr Peter Berger (Geisinger Medical Center, Danville, PA) told heartwire that this drug would be suitable for those at highest risk of another ischemic event. "Cardiologists can identify ACS patients at increased risk of suffering a recurrent ACS. For example, patients with prior episodes of ACS despite optimal medical therapy, diabetics, and those with more severe vascular disease are at a much higher risk of recurrent ACS than patients without these characteristics. Such patients might be suitable for treatment with rivaroxaban, especially if they appear to be at lower risk of bleeding."

Another question Armstrong asked was whether rivaroxaban could have been started earlier, given that one-quarter to one-third of events occur in the first few days after an ACS. And he also wondered whether three drugs were necessary. "Given the rising number of drugs we are giving these patients, is it time to start subtracting some? Perhaps we could think about replacing aspirin."

Dose One-Quarter That Used in AF

There is something else unprecedented in ATLAS 2 other than the mortality reduction, and that is the bleeding hazard.

Another outside expert, Dr Shaun Goodman (St Michael's Hospital, Toronto, ON), also said he thought the results were "impressive" and "the benefits appear to outweigh the risks, particularly with the 2.5-mg dose." Noting that this dose is just one-quarter of that used in the ROCKET-AF trial, Goodman said he thought this was the critical factor that allowed a successful result in the context of adding rivaroxaban on to dual antiplatelet therapy. But he added, "One always worries about translating these types of results into routine clinical practice, where bleeding risk (and ischemic risk) is always higher in a less selected patient population."

But commenting on the trial for heartwire , Dr Paul Gurbel (Sinai Hospital of Baltimore, MD) was more critical. "There is something else unprecedented in ATLAS 2 other than the mortality reduction, and that is the bleeding hazard. The rate of increased bleeding with prasugrel (Effient, Lilly) in TRITON and ticagrelor (Brilinta, AstraZeneca) in PLATO pales in comparison to the threefold increase in TIMI major bleeding and ICH seen with rivaroxaban in ATLAS, which in my opinion is a huge limitation."

Can It Compete With Ticagrelor?

Gurbel and Goodman also both made the point that ticagrelor also showed a mortality benefit in PLATO without such a large bleeding risk and with only two drugs, rather than the three in this trial. (Ticagrelor was given on top of aspirin alone in PLATO).

On this point, Gibson said it was inappropriate to compare ATLAS 2 with PLATO. "In PLATO, ticagrelor was given up front at the time of the event. This study is focusing more on secondary prevention, with rivaroxaban started a few days after the ACS event. That makes the reduction in mortality even more impressive, as we have missed out on the initial few days where the event rate is always high. Also, in PLATO, ticagrelor replaced clopidogrel and showed a 22% reduction in all-cause mortality. We added rivaroxaban on top of clopidogrel, which was a much tougher challenge, and achieved a 36% reduction in mortality (with the low dose)."

Secrets of Success

The ATLAS 2 results may come as a surprise to many after the APPRAISE-2 trial with another factor Xa inhibitor, apixaban, had to be stopped early because of an increased bleeding rate. This led many commentators to believe that it may not be possible to add further anticoagulant drugs onto dual antiplatelet therapy. But Gibson says he and his colleagues were encouraged by the results of the earlier ATLAS 1 trial, which showed a 30% reduction in death/MI/stroke. "Because we saw such good results in this earlier trial, we believed it would be possible to add benefit on top of dual antiplatelet therapy."

He said there were three points that had contributed to the current trial's success: selecting a lower dose than the equivalent of apixaban used in APPRAISE-2; excluding patients with previous stroke or TIA; and including mainly high-risk patients who had the most to gain from additional therapy.

The trial was supported by Johnson & Johnson and Bayer Healthcare. Gibson has received research grant support and consulting fees from Johnson & Johnson and consulting fees from Bayer, among other pharmaceutical companies. Disclosures for the coauthors are listed in the paper. Ohman has consulted for Bristol Myers Squibb, Liposcience, Merck, Pozen, Roche, Sanofi-Aventis, the Medicines Company, WebMD, and Gilead Sciences; grants from Eli Lilly and Maquet; and has served on speaker's bureaus for Boehringer Ingelheim, Gilead Science, and the Medicines Company. Goodman receives research grant support and speaker/consulting honoraria from Bayer and Johnson & Johnson and Bristol-Myers Squibb. He was involved in ROCKET AF and APPRAISE-2 as a steering-committee member. Gurbel receives research funding from and has consulted for AstraZeneca (on ticagrelor), and has sat on an advisory board for rivaroxaban.


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.