Don't Assume Newer Is Better for All -- Randomize: Lessons From FRAIL-AF

John M. Mandrola, MD; Geert-Jan Geersing, MD, PhD; Linda Joosten, MD, MSc


September 25, 2023

This transcript has been edited for clarity.

John M. Mandrola, MD: Hi, everyone. This is John Mandrola from | Medscape Cardiology. I'm here at the European Society of Cardiology meeting and I am super-excited to have the investigators of the FRAIL-AF study, which is a really neat trial. I just learned about it at the meeting.

This is a trial looking at withdrawal of vitamin K antagonists — that's warfarin, mostly, for US doctors — to non–vitamin K oral anticoagulants (NOACs), also called direct-acting oral anticoagulants (DOACs), looking at a primary endpoint of major bleeding. This is in frail patients, not typical trial patients.

I'm very excited to have our two trialists. They are very Dutch, from the Utrecht University, and I'm not going to mess up the pronunciation of their names, so they're going to introduce themselves.

Linda Joosten, MD, MSc: Hi. I'm Linda Joosten. I'm a PhD student and a medical doctor at Utrecht University in the Netherlands.

Geert-Jan Geersing, MD, PhD: My name is Geert-Jan Geersing. I'm a GP and associate professor at Utrecht University.

FRAIL-AF Key Findings

Mandrola: Linda, you were the first author of a big presentation in the main auditorium. Tell us what the results were of the FRAIL-AF trial.

Joosten: The main results were very surprising and unexpected. The FRAIL-AF study was designed as a superiority trial, with the hypothesis that switching from a vitamin K antagonist to a NOAC would lead to less bleeding. However, we saw the opposite, and that's also why the Independent Data and Safety Monitoring Board advised to stop the inclusion of patients.

We saw that switching from a vitamin K antagonist to a NOAC could lead to more bleeding, with a hazard ratio of 1.69 and a highly significant P value of .001, which means that switching from a vitamin K antagonist to a NOAC, compared with continuing a vitamin K antagonist in frail older patients, is clearly contraindicated.

Mandrola: This is really surprising because you designed a trial to show superiority of NOACs for less bleeding and you found the opposite. Did the increase in bleeding occur because of switching? I know that in my clinic it's hard to switch. Was it early on?

Geersing: No, actually, it was not. If you look at the cumulative incidence curve, you can see that the curves separate later in the trial. The switching itself went well. We were really on top of this. We were really stringent in how to switch. The curves separate later in the study, so it's not the switching. It is later in follow-up.

Mandrola: One of the reasons why I was drawn to this trial is because so many of my patients are frail older adults or they have multimorbidity. I'm thinking, when I'm seeing these patients in my clinic, that they aren't like patients that are in the trials. You enrolled special patients, I think.

Joosten: We enrolled frail older patients; we found them with the Groningen Frailty Indicator, which is a validated questionnaire that assesses frailty on several domains, such as mobility, cognition, and also the psychosocial domain of those who are really frail older patients with a loss of mobility or hearing loss.

Mandrola: I was going to ask you about the frailty index. A Groningen score of 4, say — what does that look like for a patient?

Geersing: They're all old, so they're typically above the age of 80 years. They would all have polypharmacy. That's not the only defining item. They will all have multimorbidity. The distinction between our trial patients and those included in the previous trials is that these patients have problems in multiple other domains. It's cognition, being independent of others, and unable to walk within their own houses without help. These type of patients are included in our trial.

We started this trial basically out of frustration because these patients are never included in trials. They are either excluded because they are at high risk of bleeding, so we don't want them to be in trials, or they are protected by their own physician to not go onto a scientific study because it's too burdensome or it's too difficult for them to enter these studies.

Then, in guidelines, we say we don't have evidence for frail elderly, but we also don't have evidence that the drugs work differently in these patients. Then we say, because we don't have the evidence and we never study it, oh, it's going to be the same, so why bother?

I was so frustrated about this because I was saying that this is simply not true. You cannot just extrapolate findings from studies that were done in healthier patients — which are really good trials, by the way. I have nothing against the NOAC trials. Indeed, if you are a nonfrail older patient, probably a NOAC will be better for you. That doesn't mean that it's the same for frail older patients. We found out that it's the opposite.

Dutch INR Monitoring and TTR

Mandrola: One of the factors that I've heard is that the Netherlands has amazing international normalized ratio (INR) management, and if this were done somewhere else where the INR management wasn't as good, then maybe these wouldn't have been the findings. Was INR management so good in the Netherlands? Tell me the story.

Joosten: We don't have time in therapeutic range (TTR) levels for individual patients, but we looked at TTR levels of frail patients who were visited at home at the Dutch thrombosis services that included our patients. There, we saw that TTR levels were around 65%. That's comparable, for example, with patients included in the ARISTOTLE trial, where TTR levels in the Netherlands were also around 65%. Those TTR levels were similar to those in countries like Canada, USA, UK, and Germany.

Mandrola: It's good, but it's not any better than the trials.

Geersing: No, I don't think so. You can still argue that if you had a patient on a low TTR level, then maybe it's still worthwhile to switch. We're going to look at it, probably, in the future, but we don't have the data yet available. It's important to know that these were patients who were on stable vitamin K antagonist management.

This is the domain of the study. These were frail older patients who had stable vitamin K antagonist management, had a TTR around 65%. That's good, but it's not perfect. The countries where TTR levels are even higher are the Nordic countries like Norway, Finland, and Denmark. They are really good at INR management. We are pretty good at it, but we're not that much better than the USA. I don't think it's going to be a big difference in our trial.

Mandrola: A 69% increase in bleeding — when you saw that, what did you think?

Joosten: We were very surprised. We didn't expect it.

Geersing: Yes, indeed. When you first see the data, your first reaction is, I don't believe this. There has to be something wrong. Why is this happening? Then the lines separated more and more, and the Data and Safety Monitoring Board said we needed to stop the trial because the difference is too big to just be a coincidence. It's a randomized design, so the patients were the same when they were randomized. This was the reason we started the trial, because we were uncertain. We were expecting to see a benefit for NOAC treatment but, of course, we were uncertain about this, and this is why we started the trial.

Mandrola: I suspect that you probably had colleagues telling you that it would be better to switch because this is a general feeling. This is the general feeling, at least, that we have in our community that DOACs are going to be safer, with less bleeding, because that is what we saw in the trials. I think what you've shown is that it's not the same.

Geersing: No, exactly. If you do look at the trials more in detail, the previous NOAC trials, there you can see that the benefits of NOACs were the strongest in the younger patients who were vitamin K antagonist naive when they entered the trial. In that sense, maybe it was not that unexpected because the DOAC trials already learned that the benefits of DOACs are probably biggest in younger patients without polypharmacy, without multimorbidity, and were vitamin K antagonist naive.

I think still, for those patients, DOAC treatment is a really good option and probably safer, or at least the trials showed that they were safer than vitamin K antagonists. The patients that we have in our trial were not included in the DOAC trials. They're about 10 years older, with frailty, multimorbidity — it's all these things together.

Starting OAC vs Switching

Mandrola: Do you feel that your findings inform the decision of which anticoagulant to start initially, de novo, in a patient who is frail with multimorbidity?

Geersing: That's a good question. I think it is always good to stick to the facts, and this is not what we studied. This was a switching trial. It's also really good for our colleagues around the world that our trial provides evidence for patients who are on vitamin K antagonist management. The question was, do we need to switch them — yes or no?

We didn't include new patients in this trial. I think it's really tempting to think about this, and I think the most important thing is we need to do more trials in these patients and also with the novel factor XI inhibitors that will enter the market. These newer drugs need to be studied upfront in frail elderly patients, not 10 years later. That's too late. We need to do this early on.

I think we need to do more studies in this population; this is what our study really taught us. We don't know for sure that it's the same.

Mandrola: I cannot agree more because exactly what we need is to study these things. We have a minute or two left. I want to ask you, how do you accomplish this? Often, I say we should study this and I'm told it's not easy. You all did it. You all figured out how to do this in these outpatient centers. Tell us how that came about.

Geersing: It's difficult. It's hard work, but it's about working as a team together. To start with, I think we had a really good team with Linda, the research nurses, and other colleagues. We had help from many people, including medical students. It was really hard work but it was also fun. We never had any dull moments with FRAIL-AF. Part of the study was run during COVID times, so we had to lock down our study for a period of time.

The most important thing is to keep your goal in mind, think about the fun part as well, and just work together as a team. I think that's what made us get to the results that we have today.

Mandrola: Excellent. I really am happy to speak with you. I'm so grateful for this trial. I think it's one of the most important trials from this meeting. Congratulations!

Joosten: Thank you.

Geersing: Thank you.

Follow John Mandrola on X (formerly known as Twitter)

Follow | Medscape Cardiology on X (formerly known as Twitter)

Follow Medscape on Facebook, X (formerly known as 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.