Insertable cardiac monitor yields high detection rate of AF in real-world population with cryptogenic stroke

In patients with cryptogenic stroke, an insertable cardiac monitor had a higher rate of detection of atrial fibrillation at 1 year in a real-world population than had been seen in a randomized clinical trial.

More than 70% of patients in the cohort in whom AF was found would have had their AF go undetected if monitoring had been limited to 30 days, as has traditionally been the practice.

Researchers presented their findings at the annual meeting of the American Academy of Neurology.

John D. Rogers, MD, FACC, chair of cardiology and director of cardiac pacing and tachyarrhythmia device therapy at Scripps Clinic, La Jolla, California, and colleagues analyzed the performance of the insertable cardiac monitor (Reveal LINQ, Medtronic) for detection of AF in a real-world population of 1,247 patients (mean age, 65 years; 53% men) with cryptogenic stroke. Mean follow-up was 326 days.

John D. Rogers

Rogers and colleagues quantified the AF detection rate at 1 year, compared it with that seen in the CRYSTAL AF randomized controlled trial of patients with cryptogenic stroke and analyzed duration of AF episodes and time to first detection of AF. Patients assigned insertable cardiac monitoring in CRYSTAL AF received an earlier version of the Medtronic device (Reveal XT).

Increased detection

“In CRYSTAL AF, there was a very precise definition of what cryptogenic stroke was,” Rogers said in an interview with Cardiology Today. “The present study was a look at real-world patients, a larger group, but also with what is more common in real-world practice, where not every patient has all the studies performed that lead to the diagnosis of cryptogenic stroke. In this real-world population, cryptogenic stroke was defined as what the attending physician determined it to be. Not all patients might have had transesophageal echocardiography or advanced imaging.”

Rogers and colleagues detected 1,737 episodes of AF in 192 patients during the study period, for a 1-year AF detection rate of 16.3%, which is 32% higher than the 1-year AF detection rate seen in the insertable cardiac monitoring group in CRYSTAL AF.

“It turns out we’re finding even more AF than the CRYSTAL AF study suggested,” Rogers said in an interview. “This supports the use of long-term monitors. Our current guidelines suggest that monitoring for 30 days sometime within the first 6 months after a stroke just isn’t enough. If 30-day monitoring were all that had been done in this study, over 70% of patients with AF would have been missed.”

30 days not enough

The median time of a patient’s longest detected episode of AF was 2.8 hours (interquartile range, 0.5-10.9) and the median time to detection of AF was 86 days (interquartile range, 20-173), according to the researchers.

“As more physicians learn that it’s not enough to monitor for 30 days, this will change clinical practice, and it will increase the use of these monitors to find AF in these patients,” Rogers said. “We now have a growing body of evidence ... to see that we’re missing a lot of AF if we don’t look harder.” – by Erik Swain

Reference:

Rogers JD, et al. Abstract I2.005. Presented at: American Academy of Neurology Annual Meeting; April 15-21, 2016; Vancouver.

For more information:

John D. Rogers, MD, FACC, can be reached at 10666 N. Torrey Pines Road, La Jolla, CA 92037; email: Rogers.John@scrippshealth.org.

Disclosure: The study was funded by Medtronic. Rogers reports receiving speaking fees, consultant fees and research grants from Medtronic and speaking fees and research grants from St. Jude Medical.