CHICAGO (FRONTLINE MEDICAL NEWS) – Patients with atrial fibrillation who frequently have a supratherapeutic international normalized ratio are at sharply increased risk for developing dementia, according to a large observational study.

“We postulate that the mechanism is an accumulation of microbleeds in the brain,” Dr. T. Jared Bunch said at the American Heart Association Scientific Sessions.

“In patients with hypertension, a condition that’s extremely common with atrial fibrillation, these repetitive small bleeds are preferentially in the hippocampus, where memory is stored,” added Dr. Bunch, who is medical director for heart rhythm services at the Intermountain Medical Center Heart Institute in Salt Lake City.

He presented a study of 1,031 patients with atrial fibrillation (AF) in Intermountain’s centralized anticoagulation service. All were on dual therapy with warfarin plus aspirin or, much less commonly, another antiplatelet agent. At baseline, their average age was in the early- to mid-70s, and none of the subjects had a history of stroke or any notes in their medical record suggestive of early cognitive decline. At this dedicated anticoagulation center, their INR was measured on a weekly or biweekly basis, as a result of which their average time spent in the therapeutic range of 2.0-3.0 was relatively high at nearly 70%.

The increased risk of dementia in patients with AF has previously been recognized. The association is stronger in patients under age 75 than in those who are older. But the mechanism has been unknown. Dr. Bunch and coinvestigators decided to test their hypothesis that the mechanism involves microbleeds secondary to long-term overanticoagulation by dividing the patients into three groups based upon their percentage of INR measurements above 3.0 during a mean follow-up of more than 4 and up to a maximum of 10 years: 240 patients had a supratherapeutic INR 25% of the time or more; 374 did so less than 10% of the time; and 417 had an elevated INR 10%-24% of the time.

The incidence of dementia diagnosed by a consultant neurologist during follow-up was 5.8% in the group with an INR above 3.0 at least 25% of the time, more than twice the 2.7% rate in patients with a high INR less than 10% of the time. In the middle group, the incidence of dementia was 4.1%. In a multivariate Cox regression analysis, having an INR above 3.0 on at least 25% of occasions was independently associated with a 2.59-fold increased risk of developing dementia, making it by far the most potent risk factor in their analysis.

The next step in their research will be to perform serial brain imaging and volumetric scans, Dr. Bunch said. Also, he and his coworkers are 3 years into an ongoing study looking at the incidence of dementia in AF patients on the various novel oral anticoagulants, where INR is a nonissue. Their hypothesis is the dementia risk will be lower than in patients on warfarin. Dr. Bunch has particularly high hopes for AF patients on apixaban (Eliquis) because it’s known to have a reduced risk of large bleeds, stroke, and GI bleeding; the hope is it will cause fewer cerebral microbleeds as well.

In an interview, the cardiologist said he believes his study showing an increased risk of dementia in AF patients with supratherapeutic INRs on warfarin plus antiplatelet therapy holds several important lessons for AF patients and physicians alike.

For patients, the message is don’t just start taking aspirin on your own because you’ve read it’s good for your heart or may reduce cancer risk; consult your physician.

And for physicians, it’s important to ask all patients on warfarin if they’re using aspirin; many don’t ask. Also, periodically reconsider the need for dual therapy with warfarin and aspirin.

“We find the risks of stroke and bleeding change dynamically over time, so the initial therapy for stroke prevention may not be the ideal therapy after 5-10 years,” Dr. Bunch said.

Lastly, for patients who are overanticoagulated on a substantial percentage of their INR measurements, it’s essential to consider a change in strategy. Either follow their INRs more closely and adjust warfarin dosing accordingly, or switch to one of the novel, more predictable oral anticoagulants, he concluded.

This study was funded internally by Intermountain Healthcare. Dr. Bunch reported having no financial conflicts.