Observational Study Supports Longer-Term Monitoring for A-fib After Cryptogenic Stroke

Confirming randomized trial data, a real-world study shows that monitoring patients who have had a cryptogenic stroke for up to a year with an insertable cardiac monitor (ICM) boosts detection of clinically significant A-fib episodes compared with the briefer 30-day monitoring advocated for in current guidelines. In fact, 72% of patients with A-fib were identified between 30 days and 1 year.

Results from both the current study and CRYSTAL AF support the concept that monitoring for 30 days or less is not sufficient to exclude A-fib after a cryptogenic stroke, according to John Rogers, MD (Scripps Institute, La Jolla, CA), who presented the findings on April 20, 2016, at the American Academy of Neurology annual meeting in Vancouver, Canada.

Guidelines from the American Heart Association/American Stroke Association state that for patients who have had an acute ischemic stroke or transient ischemic attack with no known cause, it is reasonable to perform prolonged monitoring—defined as about 30 days—for A-fib within 6 months of the event.

Rogers, noting that CRYSTAL AF, EMBRACE (another trial of rhythm monitoring), and the guidelines all came out at about the same time, told TCTMD that there are now enough data to support a change to the guidance indicating the need for longer-term monitoring. He said it’s uncertain whether the recommendation would be to start with 30-day monitoring or an ICM, but he added, “I can’t see how it won’t change the guidelines because the A-fib is being detected at a much higher rate.”

And that, in turn, “will help to prevent further stroke,” he said.

Brian Silver, MD (Rhode Island Hospital, Providence, RI), however, was not ready to call for a change to the guidelines.

Commenting on the study for TCTMD, Silver said that what is still needed is a trial comparing 30-day monitoring with an external device and 1-year monitoring with an ICM to see whether there is, in fact, a difference in A-fib detection. Although he said it makes sense that longer monitoring would detect more cases, he pointed out that the rate of A-fib detection seen with 30-day monitoring in the EMBRACE trial (16.1% for episodes lasting at least 30 seconds) was not very different from the rates seen after 1-year of continuous monitoring in CRYSTAL AF (12.4% for episodes lasting more than 30 seconds) and the current study (16.3% for episodes lasting at least 2 minutes).

A rate in the range of 10% to 20% is expected, Silver noted.

This study “confirms that there’s a large burden of atrial fibrillation in patients who have had a cryptogenic stroke and that we should do extended monitoring on these patients, at least 30 days,” he said. However, he added, “it remains to be determined whether doing prolonged monitoring beyond that provides benefits both in terms of detection rate and in terms of cost-effectiveness.”

Yield of Longer-Term Monitoring

About 30% of ischemic strokes are classified as cryptogenic. A-fib is one potential cause, but it might not always be detected with standard monitoring. Finding A-fib if it’s there is considered important, because it would justify switching patients from antiplatelet therapy—which is inadequate if the arrhythmia is present, Rogers said—to anticoagulant therapy.

CRYSTAL AF showed that continuous monitoring with an ICM—Reveal XT (Medtronic)—for up to a year detected more A-fib than standard care in patients who had had a cryptogenic stroke or transient ischemic attack.

The current study examined the same issue in a real-world cohort of 1,247 patients with cryptogenic stroke—as determined by the treating physicians—who received Medtronic’s next-generation Reveal LINQ monitor, which automatically transmits data from the device over a cellular network.

During follow-up, there were 1,737 A-fib episodes lasting at least 2 minutes recorded in 192 patients. The 1-year detection rate was 16.3%, which compared with just 4.6% at 30 days.

Rogers reported that the 1-year detection rate was higher in the current study than in CRYSTAL AF, although the difference did not reach statistical significance (16.3% vs 12.4%; P = 0.15).

Treatment Decisions

Silver noted that patients with cryptogenic stroke are not automatically placed on oral anticoagulation, but instead receive antiplatelet therapy. Detecting A-fib lasting at least 2 minutes can justify switching treatment to oral anticoagulation, he said.

“We know that atrial fibrillation increases the risk of stroke five-fold and antiplatelet drugs are not anywhere near as effective as anticoagulants at lowering that risk,” he said.

Whether all patients with cryptogenic stroke should be placed on oral anticoagulation as the default remains an actively investigated question, he said, pointing out that trials that are ongoing or in the planning stage are randomizing such patients to aspirin or one of the new oral anticoagulants.

If those trials show that immediate treatment with oral anticoagulation is beneficial after cryptogenic stroke, the question of how long rhythm monitoring should be used “would then become a moot point,” Silver said.

 

---

 

 

 

Related Stories:

• Long-term Cardiac Rhythm Monitoring More Likely to Detect A-fib After Cryptogenic Stroke
• PFO May Be ‘Incidental’ in One-Third of Cryptogenic Strokes
• Clinicians Urged to Enroll PFO Patients with Cryptogenic Stroke in Clinical Trials