CHICAGO -- Researchers failed to prevent cognitive impairment in people with existing atrial fibrillation (Afib) by going the route of anticoagulation to mitigate subclinical brain infarcts, according to the BRAIN-AF trial.
Stopped early for futility, BRAIN-AF suggested a similar risk of cognitive decline, stroke, or transient ischemic attack (TIA) whether Afib patients with a low risk of stroke -- a group without an existing indication for anticoagulation -- were assigned lower-dose rivaroxaban (Xarelto) or placebo (7% vs 6.4%, HR 1.10, 95% CI 0.86-1.40). The lack of between-group difference in the primary outcome was consistent across prespecified patient subgroups.
At least there was no harm detected for rivaroxaban in this setting: major bleeding reached 0.3% with rivaroxaban versus 0.8% with placebo, reported Lena Rivard, MD, MSc, of the Montreal Heart Institute, at the American Heart Association (AHA) annual meeting.
She noted that the trial had been terminated early for futility after an interim analysis; originally scheduled for 5 years, the investigators ended up with follow-up averaging 3.7 years of regular cognitive testing.
In the aftermath, it remains unclear how cognition relates to Afib -- whether one causes the other or if they simply coexist due to shared risk factors.
If the mechanism of cognitive decline in Afib were really related to microemboli going from the heart to the brain, then anticoagulation would be expected to work for prevention, commented Andrea Russo, MD, of Cooper University Health Care in Camden, New Jersey, who was not involved with the trial.
"Presence of microemboli is only one mechanism. We need to study other mechanisms, such as cerebral hypoperfusion, maybe the impact of drugs used to treat Afib," Russo said at an AHA press conference.
However, AHA session discussant Hooman Kamel, MD, of Weill Cornell Medicine/NewYork-Presbyterian in New York City, pointed out that BRAIN-AF may not be enough to exclude brain infarcts as a major mechanism of cognitive decline in Afib.
The study population was quite young and healthy, he noted, citing the average participant age being just over 53.
Notably, the trial excluded people with an indication for anticoagulation or antithrombotic therapy, prior stroke or TIA, hypertension, diabetes, congestive heart failure, or a condition associated with bleeding risk.
Rivard acknowledged that BRAIN-AF results may not be extrapolated to these other populations.
BRAIN-AF was conducted across Canada and had patients randomized at 53 sites from 2015 to 2023. Included were 1,235 Afib patients, ages 30-62, with a low risk of stroke per Canadian guidelines.
About a quarter of the cohort were women and over 95% were white. The pattern of Afib was paroxysmal in over 78% of cases, persistent in 11%, and long-standing persistent in around 11%.
The study protocol had participants randomized to rivaroxaban 15 mg or placebo. Patients who developed any of the standard indications for anticoagulation therapy were withdrawn from the study medication and switched to standard anticoagulation therapy.
As expected, there was a low incidence of stroke (0.8%) across the whole study cohort.
Among those who logged endpoint events, cognitive decline was by far the most common outcome (91.4%), with stroke (5.1%) and TIA (3.5%) less frequent.
For their cognitive testing, all patients underwent a Mini-Mental State Examination (MMSE) and Montreal Cognitive Assessment () at baseline. The MoCA was repeated yearly, while the final visit had participants undergo a last MMSE and MoCA. Cognitive decline was defined as a 2-point+ reduction in the MoCA from baseline.
This reliance on the MoCA alone for cognitive assessment was a major study limitation, the authors acknowledged. Additionally, the results are specific to low-dose rivaroxaban and may not apply to another oral anticoagulant.
"[BRAIN-AF] really should set the stage for future studies examining ... cognitive decline or dementia as an important trial endpoint in the future, in addition to our classic endpoints of things like hospitalization and stroke," Russo said.
"Emerging [Afib] treatment strategies should incorporate measures of cognition from the start," Kamel agreed.
![author['full_name']](https://clf1.medpagetoday.com/media/images/author/nicoleLou_188.jpg)
Disclosures
BRAIN-AF was funded by the Canadian Institutes of Health Research, Heart and Stroke Foundation of Canada, Canadian Stroke Prevention Intervention Network (CSPIN), and Montreal Heart Institute Foundation. Bayer provided in-kind contribution of study drugs and a grant.
Rivard disclosed research grants and speaking fees from Fonds de Recherche en Santé du Québec and CSPIN.
Russo disclosed relationships with Bayer, Boston Scientific, Medtronic, Abbott, Atricure, Biosense Webster, Biotronik, and PaceMate.
Kamel disclosed research support from the NIH, Bristol Myers Squibb, Roche; advising/endpoint adjudication committee relationships with AbbVie, Arthrosi, AstraZeneca, Boehringer-Ingelheim, Eli Lilly, Medtronic, Novo Nordisk; and ownership interest in TETMedical, Spectrum Plastics, and Ascential Technologies.
Primary Source
American Heart Association
Rivard L, et al "The BRAIN-AF trial" AHA 2024.