ARIES-HM3: Aspirin Increases Bleeding Risk Following LVAD Implantation


Results of ARIES-HM3 demonstrate use of aspirin was associated with an increased rate of major bleeding following HeartMate 3 implantation.

Mandeep Mehra, MD |  Credit: Brigham and Women's Hospital

Mandeep Mehra, MD
Credit: Brigham and Women's Hospital

New data from a trial presented at the American Heart Association Scientific Sessions 2023 suggest use of aspirin in patients with a fully magnetically levitated left ventricular assist device (LVAD) is putting patients at an unnecessary bleeding risk.

Results of the trial, which compared use of aspirin against nonuse in patients undergoing LVAD implantation with the HeartMate 3 pump and receiving a vitamin K antagonist, suggest those not taking aspirin had a rate of bleeding 33% less than those taking aspirin, which results in 47% reduction in days in the hospital among this patient population.1

"The ARIES study moves the needle forward in improving the journey of advanced heart failure patients with a marked improvement in bleeding events, healthcare resource use and cost-savings by a simple decision to avoid the use of aspirin," said Mandeep R. Mehra, MD, executive director of the Center for Advanced Heart Disease and the William Harvey Distinguished Chair at Brigham and Women's Hospital.2 "The data is so compelling that the magnitude of benefit observed in avoiding aspirin is similar to the impact of introducing a new device to the market."

An international, randomized study of 100 mg aspirin or placebo with vitamin K antagonist therapy in advanced heart failure patients with Abbott's HeartMate 3 LVAD, the trial was conducted from July 2020 through September 2022 and included 628 patients from North America, Europe, Asia, and Australia. Of the 628 who underwent randomization, 296 patients in the placebo group and 293 in the aspirin group were in the primary analysis population.1

The trial’s primary outcome was a composite endpoint, assessed for noninferiority at a margin of −10% of placebo, was survival free of a major nonsurgical hemocompatibility-related adverse events, including stroke, pump thrombosis, major bleeding, or arterial peripheral thromboembolism, at 12 months.1

Initial analysis revealed more patients were alive and free of hemocompatibility events at 12 months in the placebo group (68%) than among the aspirin (74%) group. When assessing noninferiority of placebo, results indicated the absolute between-group difference in improvement in event-free survival with placebo was 6.0% (lower 1-sided 97.5% Confidence Interval, −1.6%; P < .001). Further analysis suggested aspirin avoidance was associated with reduced nonsurgical bleeding events (relative risk, 0.66 [95% confidence limit, 0.51 to 0.85]; P = .002) with no increase in stroke or other thromboembolic events, which investigators pointed out was consistent across subgroups.1

In their release announcing results, Abbott pointed out labeling changes related to the study's anticoagulation regimen have not been approved by the FDA at this time.2 For more on the results of the trial, check out our Q&A with Mehra from the floor at AHA 2023.

HCPLive Cardiology: Can you provide a main overview of this trial and what it adds to our understanding of management in patients receiving LVAD?

Mehra: For about 4 decades, we have been trying to create an option for treatment of patients who are pharmacologically nonresponsive with heart failure and left ventricular assist devices have been that answer, but they have been an imperfect answer for a very long time. Much of the gains that we have seen is the notion of ventricular assist device therapy has moved from rescue therapy— initially we would use themin patients that were dying and now they are moved from rescue to actual support—to chronic support and durable support.

The problem with that is that these patients continue to suffer a lot of complications. We call this constellation of complications, hemocompatibility-related complications. These are basically 3 complications. One is strokes. the second is pump thrombosis, which basically requires reoperation and it's a very big deal. Then, the third is bleeding—predominantly, gastrointestinal bleeding.

Now, this is not a traditional form of bleeding, it's a very unique form of bleeding that is due to angiodysplasia, and arteriovenous malformations in the gut. It's due to the unnatural physiology that's imparted by left ventricular assist devices.

We have actually shown that with the latest generation pump, the HeartMate 3, that pump thrombosis has nearly been eliminated. Strokes are down now to less than 2 or 3% per year and might in fact be ambiently at the same rate as what you and I would encounter in an advanced heart failure population, as long as the acute operation is successful without a stroke. However, gastrointestinal bleeding has continued to be one of the top 3 reasons for rehospitalization of patients and, often, they require to be admitted to the hospital blood transfusions and the cost of that hospitalization is quite high.

If you look at it in the United States federal payer construct, which is Medicare, the cost of that of a single hospitalization for a gastrointestinal bleed is about $13,500 for Medicare. Now, we know Medicare always under pays the cost. So, the cost in most hospital systems is then transferred to commercial payers and you can imagine that the actual cost is somewhere in the $20,000 plus range for every hospitalization.

Why is that? Well, patients come and get admitted. They then get blood transfusions; they're constantly getting blood tests. A gastroenterologist may come and do an endoscopy and often the yield is less than 50%. Very often people will do multiple endoscopies. So, that is why the costs of this are huge to the healthcare system, in addition to the morbidity.

We have we have been struggling with what to do with bleeds and there has been no real good treatment. We also lack good diagnostics.

So, we took a step backwards and said, 'Wait a minute, we have a fantastic new pump. It was an engineered reasonably well. Are we doing the right thing with keeping people on aspirin and Vitamin K antagonists together?'

It turns out that ARIES-HM3 answered that question extremely convincingly. We see that even what is so called baby aspirin is actually a poison in these patients and causes a substantial number of bleeds. To the point that pulling it off reduces hospitalizations by 47%, the cost of hospitalizations by 41%, and the actual gastrointestinal bleeds by 40%. That magnitude of effect is something that you normally would see if you introduced a new therapy. So, it's very rare in medicine to just take something away and realize that you've been doing the wrong thing for a long time.

HCPLive Cardiology: Can you talk about the comfortability that that should give clinicians when sort of moving forward without aspirin given the consistency observed across patient subgroups?

Mehra: So, that is the pivotal finding of this trial, right? We actually didn't have a situation where we have to rethink this. It's generalizable across the board, anyone coming with a HeartMate 3 should basically not be on aspirin. That's essentially what it is.

What we're also saying is that you can be very well assured that there is no situation in which you're bringing the patient into the lab that requires your attentiveness to put them on aspirin, right? Now, what does that actually mean clinically? What that clinically means is that when you look at these so-called subgroups, and you find that removal of aspirin works in all of them, it actually is telling you that the safety that is imparted is potentially from the vitamin K antagonists because there are subgroups that vary in thrombotic risk. When you see such a consistent effect, that means that the background therapy with a vitamin K antagonist, which was similar in both groups, is what is providing at least some reasonable stability, but it also tells you that may be the next thing for us to look at a bit more carefully.

As you know, our time in therapeutic range with warfarin-like drugs, like Vitamin K antagonists, was only 55 to 58% in this trial between the two groups. So, it is the next thing that we need to start focusing on: Whether we need a much better vitamin K antagonist for these patients or whether we need a vitamin K antagonist at all. I would think that we're going to end up needing something, but that is just not aspirin and that's what the trial is showing us.

HCPLive Cardiology: Can you talk a bit about how much the field has advanced in recent years and what the next few years might tell us about antithrombotic therapy?

Mehra: I think that the advances are huge, actually. At least for high-risk situations, like putting metal into your chest cavity with a left ventricular assist device. I think the opportunity of these new agents is really quite exciting. So, if you think about where the field is heading: we need to figure out a better vitamin K antagonist. That could be one. Or maybe we need to actually test DOACs versus vitamin K antagonists in these patients with LVAD. Many people will try away from that because DOACs tend not to work as well in very high-risk situations, like if someone has a left ventricular clot etc, but I am quite excited by these factor XI inhibitor drugs because they tend to offer selectivity in reducing thrombosis but not causing bleeds.

Even if you look at ARIES-HM3, with the massive reduction in bleeds that we saw, well as massive as it is it moves from 42% at 2 years to 30%. So, it's a significant drop in the incidence of bleeding, but the frequency or rates of bleeding is still 30% and that rate of bleeding is significant.

So, we still have a way to go and we need to keep on pushing the mantle on looking at that these therapies. Basically, what we've said is instead of 2, do you need 1? Now the question is how good should that 1 be or which 1 should it be. That's where we have to go.


  1. Mehra MR, Netuka I, Uriel N, et al. Aspirin and Hemocompatibility Events With a Left Ventricular Assist Device in Advanced Heart Failure: The ARIES-HM3 Randomized Clinical Trial. JAMA. Published online November 11, 2023. doi:10.1001/jama.2023.23204
  2. Late-breaking data show an aspirin-free medication regimen benefits people with Abbott’s Heartmate 3TM Heart pump. Abbott MediaRoom. November 13, 2023. Accessed November 13, 2023.
Recent Videos
1 KOL is featured in this Insights series.
Charles C. Wykoff, MD, PhD: Interim Analysis on Ixo-Vec Gene Therapy for nAMD | Image Credit: Retina Consultants of Texas
Sunir J. Garg, MD: Pegcetacoplan Preserves Visual Function on Microperimetry | Image Credit: Wills Eye Hospital
Edward H. Wood, MD: Pharmacodynamics of Subretinal RGX-314 for Wet AMD | Image Credit: Austin Retina Associates
Dilsher Dhoot, MD: OTX-TKI for NPDR in Interim Phase 1 HELIOS Results  | Image Credit: LinkedIn
Katherine Talcott, MD: Baseline EZ Integrity Features Predict GA Progression | Image Credit: LinkedIn
Veeral Sheth, MD: Assessment of EYP-1901 Supplemental Injection Use in Wet AMD | Image Credit: University Retina
Discussing Post-Hoc Data on Ruxolitinib for Nonsegmental Vitiligo, with David Rosmarin, MD
1 KOL is featured in this series.
1 KOL is featured in this series.
© 2024 MJH Life Sciences

All rights reserved.