Journal Wrap | Detecting Underlying AF in Cryptogenic Stroke Patients

Although a leading cause of recurrent stroke, atrial fibrillation (AF) is often asymptomatic and likely to go untreated in the routine care of patients with stroke or transient ischemic attack (TIA). While earlier detection of AF through cardiac monitoring in these patients seems to have some value in preventing recurrent strokes, guidelines for stroke management do not endorse any specific strategies for detecting AF in patients with a new stroke. Recently, two studies in The New England Journal of Medicine investigated the potential benefit of prolonged cardiac monitoring in improving the detection and subsequent treatment of "covert" AF.


In the first report, the EMBRACE (30-Day Cardiac Event Monitor Belt for Recording Atrial Fibrillation after a Cerebral Ischemic Event) investigators compared electrocardiography (ECG) monitoring for 30 days with an event-triggered recorder (intervention group) to conventional 24-hour monitoring (control). The study randomized 572 patients without known AF who had experienced a cryptogenic ischemic stroke or TIA within the previous 6 months; 286 were included in the intervention group analysis and 277 in the control group (one patient withdrew immediately after randomization and no data were recorded). The primary outcome was detection clinically or via the study monitors of one or more episodes of ECG-documented AF or atrial flutter lasting ≥30 seconds within 90 days after randomization.

Prolonged noninvasive ambulatory ECG monitoring improved the detection of AF to a factor of more than five (TABLE):

  • The primary endpoint was detected in significantly higher rates of patients in the intervention group (16.1% vs. 3.2%; p < 0.001).
  • Episodes of AF lasting longer than 2.5 minutes (a secondary outcome) were seen in 28 patients (9.9%) in the intervention group and seven (2.5%) in the control group.

TABLE. EMBRACE:  Detection of AF in the Two Monitoring Groups

In another secondary outcome measure, intervention patients with prolonged monitoring were nearly twice as likely to be prescribed oral anticoagulant treatment at 90 days than controls (52 of 280 patients [18.6%] vs. 31 of 279 [11.1%]; p = 0.01).


In an analysis of the CRYSTAL-AF (Cryptogenic Stroke and Underlying AF) trial, Sanna et al. looked at the efficacy of a nontraditional monitoring technique using a smaller, insertable cardiac monitor or ICM (Reveal XT; Medtronic, Inc.) compared to conventional 24-hour monitoring for detecting subclinical, asymptomatic AF in stroke patients. This randomized, controlled study included 441 patients who had experienced a cryptogenic stroke within 90 days and had no evidence of AF in the 24 hours prior to randomization.

By 6 months, AF had been detected in 19 (8.9%) patients in the ICM group versus three patients (1.4%) in the control group, translating to a more than 6-fold increase in AF detection in the longer-term monitoring group (HR = 6.4; 95% CI 1.9-21.7; p < 0.001). By 12 months, that difference had increased to 7.3 times greater: 29 patients (12.4%) in the ICM group and four (2.0%) in the control group (HR = 7.3; 95% CI 2.6-20.9; p < 0.001).

In an editorial accompanying both studies, Hooman Karmel, MD, posed the question as to why prolonged rhythm monitoring has not been adopted as a standard test in cryptogenic stroke patients. The answer: Unlike these two trials, most studies have lacked control groups, leaving it unclear whether monitoring actually improved diagnosis versus routine follow-up, and the fact that subclinical AF often becomes clinically apparent AF has hindered this progress. But, importantly, he adds, "the two randomized trials show that prolonged rhythm monitoring identifies atrial fibrillation that would not have otherwise declared itself."

Missing a diagnosis of AF after stroke can have significant consequences: "Even a few minutes of this subclinical form of atrial fibrillation increases the risk of stroke... and patients without recognized atrial fibrillation typically receive antiplatelet therapy after stroke; therefore, a failure to diagnose atrial fibrillation as the cause of stroke may result in suboptimal antithrombotic therapy." Given the weight of current evidence, including the studies headed by Drs. Gladstone and Sanna, prolonged cardiac rhythm monitoring to thoroughly rule out subclinical AF and inform subsequent therapy decisions may be a protocol whose time has come for guideline recognition.

Gladstone DJ, Spring M, Dorian P, et al. N Engl J Med. 2014;370:2467-77.
Karmel H. N Engl J Med. 2014;370:2532-3.
Sanna T, Diener H-C, Passman RS, et al. N Engl J Med. 2014;370:2478-86.

< Back to Listings