Aspirin Recs Go Topsy-Turvy in New Guidelines: Now What?

Michael D. Miedema, MD, MPH; J. William McEvoy, MB BCh, MHS


October 16, 2019

Editorial Collaboration

Medscape &

This transcript has been edited for clarity.

J. William McEvoy, MB BCh, MHS: Hello. My name is Bill McEvoy. I'm a preventive cardiologist currently based at the National Institute for Prevention and Cardiovascular Health at the National University of Ireland in Galway, Ireland. I am also an adjunct professor at Johns Hopkins University in Baltimore, Maryland.

Michael D. Miedema, MD, MPH: Hello. I'm Mike Miedema. I'm director of cardiovascular prevention at the Minneapolis Heart Institute in Minneapolis, Minnesota. Dr McEvoy, I'd say "Good morning," but I believe that in Ireland now I should say "Good afternoon."

Recent Guideline Change for Aspirin

McEvoy: Yes, it's a beautiful sunny afternoon here in Ireland. We were [authors] on the 2019 American College of Cardiology/American Heart Association (ACC/AHA) Guideline on the Primary Prevention of Cardiovascular Disease.[1] We are both well versed in the evidence around aspirin, which turned out to be one of the more interesting and novel aspects to the guideline in terms of recent data that were published in 2018. The 2019 recommendations from ACC/AHA downgrade the level of recommendation for aspirin in primary prevention.

Do you think the guideline went far enough in its recommendation or went too far? The European guidelines have said for many years that aspirin should not be provided for the primary prevention of cardiovascular disease (CVD), and the change in the 2019 US guideline was a move in that direction. Was the move far enough, based on the evidence, in your opinion?

Miedema: It's a controversial and important topic and it's very relevant to most middle-aged adults, so it's gotten a lot of attention. Three recent large and well-done trials[2,3,4] clearly showed that the benefit for aspirin was quite small, if any. To scale it back was completely appropriate. Sometimes we like to make these yes-or-no answers, when in fact, nuance is really important here. The trials did show a small benefit, so I'm not sure it's an appropriate recommendation to say that we should never use aspirin for primary prevention.

McEvoy: I agree, Mike. That is basically what we said in our conversations amongst the guideline writing committee. When you contrast the prior US guidelines on aspirin for primary prevention with the most updated one from 2019, in many respects it tells the story of how the evidence has changed. The benefit for aspirin appears to be less pronounced now in high-risk primary prevention. What do you see as the main drivers of apparent change in efficacy for aspirin in more modern studies compared with the older studies?

Why Does Aspirin Seem Less Effective Than Before?

Miedema: That is a very important question. Sometimes in medicine it's better to be first than to be best. And I think for aspirin, it was first. When the first trials were done, there were not a lot of other treatments available for a primary prevention of CVD. It was an era where there were higher rates of smoking, more trans fats in the diet, and other risk factors that were more prevalent, so in these initial trials, there were a lot more events. In order to prevent events you first have to have events. In a trial with not a lot of other treatments, aspirin looked beneficial. When we repeat the trials in the modern era, where there is a lot more statin use, better blood pressure control, and less smoking, there are a lot fewer events, so the benefit of aspirin has gotten smaller from an absolute perspective. That is one of the main changes, but there are a few other things that are relevant as well.

The more recent trials have looked at a composite outcome of heart attacks, stroke, and cardiovascular death, whereas the initial aspirin study from the Physicians' Health Study in the New England Journal of Medicine[5] just looked at myocardial infarction (MI) as the primary outcome. Some people die from CVD that is not an atherosclerotic or thrombotic event—events that aspirin could not necessarily prevent—so when you add in cardiovascular death, you are diluting the benefit of aspirin a little bit.

In these recent long-term trials, there is a fair amount of withdrawal over time, so by the time you get to the end of the study, almost a third of the people have withdrawn. Those people are factored into the analysis, even though they withdrew, so it dilutes the benefit of aspirin. In fact, the per-protocol analysis of ARRIVE[2] found that there was a significant benefit for people who stayed on aspirin for the duration of the trial versus people who stayed on placebo for the duration of the trial. Event rates were still small, but there was a benefit. It gets back to this nuance where I do think there is a small benefit, but we just have to compare it to the bleeding risk.

Balance of Risk and Benefit

McEvoy: That is fair. That brings up the comparison, or the balancing between the risk and the benefit. I think both of us would agree that there is compelling evidence that aspirin can reduce nonfatal MI, maybe less so now than it could before, for the reasons you outlined. But an MI is a significant diagnosis. I can certainly think of young patients in their late 40s and 50s with a big MI, and they now have lifelong consequences of the MI in terms of ischemic cardiomyopathy or downstream issues. I'm certainly very sensitive to a therapy that can reduce nonfatal MI. It can be a challenge when you balance that with bleeding risk, because the bleeding risk with aspirin is quite heterogeneous, as in fairness is the [magnitude of] reduction in MI, I suppose. But certainly, most [aspirin] bleeds are quite easy to treat; most are minor. The proportion of people who have major bleeding or an intracranial bleed is a small minority. How do you think patients and doctors should try to balance these competing risks and benefits when, in some respects, they are not quite comparing apples with apples? It's a little bit of apples with oranges.

Miedema: I agree wholeheartedly. It's tough to quantify what exactly the impact of a gastrointestinal bleed is compared with the impact of a heart attack. That is why it's important that the guidelines use the age cutoff. In ASPREE,[3] a trial that looked at older individuals and aspirin, the rate of major bleeding events over the 5 years was 2.5%. That is a relatively high bleeding risk in an older population, whereas in ARRIVE, which was a younger population, the bleeding risk was less than 1%, and the majority of those bleeds were minor. The guidelines say that for people above age 70, you should be really cautious with any use of aspirin for primary prevention, whereas for people 40-69, there is a little bit more favorable benefit-to-risk ratio. The bleeding risk in the middle-aged adult is relatively small. So if there is elevated cardiovascular risk, we should have a discussion. [Aspirin for primary prevention] as a 2B recommendation—meaning, don't do it for sure but at least talk about it and consider it—is probably the best recommendation to give.

McEvoy: I think you are right. Those premature MIs are going to happen, of course, in the younger adults as well, and they have the most to lose from premature MIs in terms of those harms and benefits, as you've outlined.

Using Risk to Allocate Aspirin Use

McEvoy: Maybe we can use this forum to tease out the issue around the use of risk to allocate aspirin among a population under the age of 70. This is something I still struggle with and certainly something I see discussed when the guidelines are interpreted by clinicians. Prior US guidelines would have used the threshold of an over 10% 10-year risk for CVD, which has a lot of logic to it. But unfortunately it's not particularly well justified in the evidence. What do you think? Is there practical advice for a physician who is struggling with the idea of using risk but not having a go-to cut-point with a particular risk score?

Miedema: If you look at the past decade of prevention, one of the main themes is the importance of absolute risk. Previous guidelines were more numbers based. If your cholesterol is high, then maybe we should treat it. If it's not, it's probably okay. But now we're saying not "What is your cholesterol?" but instead "What is your overall level of risk? And if your risk is high, you should think about lowering your cholesterol." This risk-based approach is clearly evidence based, but it's also shown us that we're not as good at calculating risk as we'd like to be. The estimated risk in these middle-aged adults in the ARRIVE trial was 17% over 10 years but their event rate over 5 years was down around 4%, so their actual 10-year rate was probably around 8%. That is why using a strict cut-off from a risk standpoint probably isn't the right way to go about it. We should be open with the patient, saying, "There is some uncertainty here. That being said, we think you are at elevated risk, which is something to consider." It should be a shared decision-making process.

Shared Decision-Making Process

McEvoy: That is what the guidelines settled on. It can be a little bit difficult for providers when they are not giving prescriptive advice, but the fact is that the evidence is not there for a certain cut-point. To add to that, the risk of bleeding tends to track with your CVD estimate as well. Those who are high risk by atherosclerotic CVD risk scores (perhaps because of the influence of age on those scores) are also the patients that tend to be at higher risk for bleeding, so it's kind of hard to disentangle those two. Hopefully we'll have better methods to triage aspirin down the road. Obviously, the guidelines do put some perspective on the idea of risk modifiers or risk enhancers, like a family history of premature MI or presence of subclinical atherosclerosis. Hopefully there is enough in the guideline for primary care physicians and general cardiologists to use to personalize the decision with the patient. Obviously, this somewhat depends on the patient's own perspective on whether they prefer to avoid an MI or a bleed. Data in this area on anticoagulation (not so much for aspirin) suggest that many patients would prefer to avoid an MI far more than they would a bleed. In some respects, there is almost a two or three times higher preference for the avoidance of an MI. When you factor that in, aspirin starts to look a little bit better. Obviously, that is something that you have to take on a case-by-case basis with a patient, which isn't the easiest thing to do. But certainly that is something that we need to do now more and more, as emphasized in the guidelines—not just for aspirin, but also in the blood pressure and cholesterol guidelines.

Miedema: I agree. It's important to empower the patient to make the decision. It's been interesting having this discussion in clinic because a lot of people view aspirin as kind of a safety mechanism and say, "I take an aspirin per day just to be safe." I've had a fair number of patients offer some resistance after my telling them that it probably does not have the benefit that they think and it may be reasonable to stop it. It's the very opposite of a discussion about a statin. Most patients worry about the risks of statins, yet the clinical trial data are pretty compelling that aspirin actually has more risk than statins. The risk is small but it's there, and it's very definable in the trials; whereas in the trials for statins, we don't seem to see a lot of risk. Having those discussions with the patient is important because sometimes they are misinformed as far as what the evidence actually shows.

McEvoy: That is a wonderful point and I could not agree with you more. Statins have very well-described benefits, but the myalgia (which we all see) was less evident in the trials, so what do we make of all that? That is a conversation for another day.

Thanks to everybody for joining us on Medscape Cardiology. We hope you were informed and educated by our conversation. Obviously, there are no right or wrong answers and this has to be taken on a case-by-case basis. Hopefully our conversation will help you inform conversations with your patients. I'll leave it to Mike to finalize our concluding remarks.

Miedema: Thanks, Dr McEvoy. It was a pleasure to be here. I hope you found this to be of some benefit. Have a great day.

Follow Medscape on Facebook, Twitter, Instagram, and YouTube


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.