The Most Important Study From ESC: FRAIL-AF

John M. Mandrola, MD


September 01, 2023

One of the hardest tasks of a clinician is applying evidence from trials to the person in your office. At the European Society of Cardiology congress, the surprising and unexpected results of the FRAIL-AF trial confirm the massive challenge of evidence translation.

FRAIL-AF investigators set out to study the question of whether frail, elderly patients with atrial fibrillation who were doing well with vitamin K antagonists (VKA) should be switched to direct-acting oral anticoagulants (DOAC).

Senior author Geert-Jan Geersing, MD, PhD, from the University Medical Center Utrecht, Utrecht, the Netherlands, told me that frustration led him to design this study. He was frustrated that colleagues assumed that evidence in nonfrail patients can always be translated to frail patients. 

Geersing offered two reasons why common wisdom may be wrong. First was that the large DOAC vs warfarin trials included few elderly patients with frailty. Second, first author Linda Joosten, MD, made it clear in her presentation that frailty is a lot more than aging. It is a clinical syndrome, which entails a "high burden of comorbidities, dependency on others, and a reduced ability to resist stressors."

The FRAIL-AF Trial

The investigators recruited elderly, frail patients with fibrillation who were treated with vitamin K antagonists and had stable international normalized ratios from outpatient clinics throughout the Netherlands. They screened about 2600 patients and enrolled nearly 1400. Most were excluded due to not being frail.

Half the group was randomized to switching to a DOAC — drug choice was left to the treating clinician — and the other half was left on VKA. Patients were 83 years of age on average with a mean CHA2DS2-VASc score of 4. All four classes of DOAC were used in the switching arm.

The primary endpoint was major or clinically relevant nonmajor bleeding, whichever came first, accounting for death as a competing risk. Follow-up was 1 year.

The Results for Switching to DOAC vs VKA

Joosten started her presentation with this: "The results turned out to be different than we expected." The authors designed the trial with the idea that switching to DOACs would be superior in safety to remaining on VKAs.

But the trial was halted after an interim analysis found a rate of major bleeding in the switching arm of 15.3% vs 9.4% in the arm staying on VKA. The hazard ratio was 1.69 (95% CI, 1.23-2.32; P = .0012).

The Kaplan-Meier event curves reveal that the excess risk of bleeding occurred after 100 days and increased with time. This argued against an early effect from transitioning the drugs.

An analysis looking at specific DOAC drugs revealed similar hazards for the two most common ones used — apixaban and rivaroxaban.

Thrombotic events were a secondary endpoint and were low in absolute numbers, 2.4% vs 2.0%, for remaining on VKA and switching to DOAC, respectively (HR 1.26; 95% CI, 0.60-2.61).

The time in therapeutic range in FRAIL-AF was similar to that in the seminal DOAC trials.


Three reasons lead me to choose FRAIL-AF as the most important study from this year’s ESC.

First is the specific lesson about switching drugs. Note that FRAIL-AF did not address the question of starting anticoagulation. The trial results show that if you have a frail older patient who is doing well on VKA, don’t change to a DOAC. That is important to know, but it is not what gives this study its heft.

The second reason centers on the investigators choice to do this trial. Geersing had a feeling that common wisdom was wrong. He did not try to persuade colleagues with anecdote or plausibility or meta-analyses of observational studies. He set out to answer a question in the correct way — with a randomized trial.

This is the path forward in medicine. I’ve often heard proponents of observational research declare that many topics in medicine cannot be studied with trials. I could hear people arguing that it’s not feasible to study mostly home-bound, elderly frail patients. And the fact that there exist so few trials in this space would support that argument.

But the FRAIL-AF authors showed that it is possible. This is the kind of science that medicine should celebrate. There were no soft endpoints, financial conflicts, or spin. If medical science had science as its incentive, rather than attention, FRAIL-AF easily wins top honors.

The third reason FRAIL-AF is so important is that it teaches us the humility required in translating evidence in our clinics. I like to say evidence is what separates doctors from palm readers. But using this evidence requires thinking hard about how average effects in trial environments apply to our patient.

Yes, of course, there is clear evidence from tens of thousands of patients in the DOAC vs warfarin trials, that, for those patients, on average, DOACs compare favorably with VKA. The average age of patients in these trials was 70-73 years; the average age in FRAIL-AF was 83 years. And that is just age. A substudy of the ENGAGE AF-TIMI 48 trial found that only 360 of more than 20,000 patients in the trial had severe frailty.

FRAIL-AF clearly shows how cautious we should be in applying evidence gathered in younger, healthier patients to older, more vulnerable patients. That lesson extends to nearly every common therapy in medicine today. It also casts great doubt on the soft-thinking idea of using evidence from trials to derive quality metrics. As if the nuance of evidence translation can be captured in an electronic health record.

The skillful use of evidence will be one of the main challenges of the next generation of clinicians. Due to advances in medical science, more patients will live long enough to become frail. And the so-called "guideline-directed" therapies may not apply to them.

Joosten, Geersing, and the FRAIL-AF team have taught us specific lessons about anticoagulation, but their greatest contribution has been to demonstrate the value of humility in science and the practice of evidence-based medicine.

If you treat patients, no trial at this meeting is more important.

John Mandrola practices cardiac electrophysiology in Louisville, Kentucky, and is a writer and podcaster for Medscape. He espouses a conservative approach to medical practice. He participates in clinical research and writes often about the state of medical evidence. Follow John Mandrola on Twitter.

Follow | Medscape Cardiology on  X (formerly Twitter).

For more news, 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.