Rivaroxaban Cuts Major Stroke Risk in COMMANDER HF Post Hoc Analysis

The main trial, looking at HFrEF patients with CAD but no A-fib, was neutral, but investigators say the stroke data are compelling.

Rivaroxaban Cuts Major Stroke Risk in COMMANDER HF Post Hoc Analysis

ATHENS, Greece—Investigators digging deeper into the stroke results from COMMANDER HF have found that low-dose rivaroxaban (Xarelto; Bayer/Janssen) appears to “markedly reduce” the risk of stroke and TIA in patients with worsening heart failure with reduced ejection fraction (HFrEF) who are in sinus rhythm but have preexisting coronary artery disease. The benefit, they stress, was seen in patients with the highest risk for serious strokes, as identified by traditional risk scores.

Muthiah Vaduganathan, MD (Brigham and Women’s Hospital, Boston, MA), who presented the analysis as a late breaker here at the European Society of Cardiology (ESC) 2019 Heart Failure Congress, acknowledged that the study is post hoc but said it identifies a “true signal” of risk for one of the most feared complications in this group of patients.

“There is an ongoing need to search for risk-reduction strategies for stroke and I think that’s what is set up by this analysis,” he told TCTMD.

The main COMMANDER HF results, reported at the 2018 ESC Congress, showed that low-dose rivaroxaban, given at a dose of 2.5 mg twice daily, was no better than placebo at reducing the rate of all-cause death, MI, or stroke. About three-quarters of those primary endpoint events, however, were all-cause deaths, many of which were due to worsening heart failure, such that any signal of benefit with rivaroxaban on other endpoints, including stroke, were eclipsed.

At the American Heart Association 2018 Scientific Sessions, COMMANDER HF investigators reported results of a post hoc analysis looking specifically at thromboembolic events. Here, compared with placebo, rivaroxaban significantly lowered a composite of thromboembolic events. In this analysis, ischemic strokes seemed to be a key driver of the benefit and prompted the analysis that Vaduganathan presented here today.

Fatal and Disabling Strokes

In all, 150 first neurological events occurred in the more than 5,000 patients in the trial: 23 events were TIAs and the remainder were strokes, including 31% that were fatal (modified Rankin scale [mRS] score of 6), 17% that were disabling (mRS 3-6), and 51% of which were nondisabling (mRS 0-2). Stroke rates were at their highest in the first 6 months after an episode of worsening heart failure, but persisted throughout the year such that the cumulative risk rose steadily throughout the first 12 months and reached 2% by the end of the year. When the CHA2DS2-VASc score, which is well-validated in atrial fibrillation patients, was applied to this group of patients in normal sinus rhythm, the median score was 4 and each stepwise increase in the score after 5 was associated with a higher risk of stroke and TIA.

For the primary analysis, rivaroxaban reduced the risk of a first stroke or TIA by 31% compared with placebo, after adjusting for other variables (HR 0.69; 95% CI 0.50-0.95). Broken down by stroke type, investigators saw a significant reduction in the risk of ischemic stroke with rivaroxaban versus placebo (0.86 vs 1.34; HR 0.64; 95% CI: 0.43-0.95), but no significant differences for hemorrhagic stroke or TIA.

For the cohort as a whole, the number needed to treat to prevent one neurological event (all-cause stroke or TIA) was 164 per year, but this was reduced to 96 when the analysis was restricted to patients with CHA2DS2-VASc scores greater than 4.

Of note, no significant differences were seen in rates of the bleeding endpoint used in this analysis, which was “fatal bleeding or bleeding into a critical space.”

Questions and Answers

“The take-home messages from this COMMANDER HF subanalysis is that patients with heart failure with reduced ejection fraction, CAD, and sinus rhythm face risks of stroke and TIA that approach that observed in patients with chronic HF and atrial fibrillation,” Vaduganathan observed. “This risk peaks just prior to 6 months but persists throughout the observation period, and nearly half of the first primary neurological events were either disabling or fatal, and survivors face ongoing risks of major adverse cardiovascular events.”

Vaduganathan also emphasized the fact that the benefits of rivaroxaban did not appear to be offset by “an appreciable increase in major bleeding,” something that has stymied other heart failure/anticoagulation trials, although he noted that minor bleeds still occurred. Importantly, he added, “select populations with HFrEF and sinus rhythm may be identified by use of traditional risk scores, a strategy that warrants further investigation.”

John McMurray, MD (University of Glasgow, Scotland), discussing the results following Vaduganathan’s presentation, stressed the severity of the strokes documented in the study. 

“Can you think of anything more devastating than having heart failure and then having that complicated by a stroke, especially since the strokes that occur, even in patients with sinus rhythm, tend to be very disabling? Clearly this is something we should be interested in as a society,” said McMurray.

McMurray offered a few hypotheses for why this trial had shown a benefit of stroke reduction without an uptick in major bleeding, as seen in other HF trials. For one, the risk of stroke and TIA “was unusually high,” he noted, “and I’m not sure why that is—was it the population? Was it the underlying coronary disease? Or was it because strokes were not adjudicated?”

McMurray also noted that the definition of bleeding used in the analysis was “very conservative,” pointing out that similar studies have more commonly used major bleeding.

“But I do agree,” McMurray said, “with the COMMANDER HF investigators that what we all need to do is to think of better ways to identify high-risk patients and to target therapies to those patients.”

The fact that an already validated risk-score was able to help predict the risk of stroke is an important contribution, he concluded, saying, “That’s where we need to go: to see how we can safely target anticoagulation to prevent this complication.”

Speaking with TCTMD following the session, senior investigator for COMMANDER HF, Mandeep R. Mehra, MD (Brigham and Women’s Hospital), stressed that the findings warrant a trial specifically addressing patients at the highest risk for disabling and fatal strokes.

“I think we have enough information from this analysis that can start to inform the next trials,” he said. “And in fact, last night at an investigators meeting, we all agreed that the first thing we’re going to do as soon as we finish this meeting is open a dialogue for a targeted trial in stroke.”

In the meantime, he says, the findings have prompted him to calculate CHA2DS2-VASc scores in patients he sees in clinical practice who are similar to those enrolled in COMMANDER HF. If the score is high, he said, “I’m going to be super vigilant in looking for another indication to anticoagulated that patient. So I would go looking for intermittent A-fib, or, if the patient has a prevailing indication for anticoagulation such as prior stroke, etc, instead of just using antiplatelet therapy in that patient I would use that as an indication to go ahead [with low-dose rivaroxaban], because a prior history of stroke is a very strong harbinger of subsequent strokes.”

Shelley Wood is Managing Editor of TCTMD and the Editorial Director at CRF. She did her undergraduate degree at McGill…

Read Full Bio
  • Vaduganathan M. Effect of rivaroxaban on stroke or transient ischemic attack in patients with heart failure, coronary artery disease, and sinus rhythm: the COMMANDER HF trial. Presented at: European Society of Cardiology 2019 Heart Failure Congress. May 26, 2019. Athens, Greece.

  • Vaduganathan reports receiving an NIH/NCATS KL2/Catalyst Medical Research Investigator Training Award, and serving on the advisory boards of Amgen, AstraZeneca, Bayer AG, Baxter healthcare, and on clinical endpoint committees for Novartis and the NIH/NHLBI.