Paroxysmal Atrial Fibrillation: An Underdiagnosed Cause of Stroke?

Abstract & Commentary

By Dana Leifer, MD Associate Professor, Neurology, Weill Medical College, Cornell University. Dr. Leifer reports no financial relationship relevant to this field of study.

Synopsis: Long-term monitoring is likely to detect atrial fibrillation in stroke patients with frequent atrial premature beats.

Source: Wallmann D, et al. Frequent atrial premature beats predict paroxysmal atrial fibrillation in stroke patients: An opportunity for a new diagnostic strategy. Stroke 2007;38:2292-2294.

Approximately 30% of ischemic strokes do not have a definite etiology (cryptogenic) even after diagnostic evaluations that are considered comprehensive by current standards. It is often thought that many of these strokes are embolic in origin. One possible explanation for some of these cryptogenic strokes is paroxysmal atrial fibrillation (AF). Paroxysmal AF may not be detected by conventional cardiac monitoring, which is usually limited to a 24-hour Holter monitor or to cardiac telemetry while the patient is hospitalized.

One recent study found that new AF was diagnosed on the admission electrocardiogram in 11 of 149 patients with an acute stroke or transient ischemic attack (TIA) and without a prior history of AF. AF was identified by 24-hour Holter monitors in an additional 7 patients and by 7-day event loop recordings in another 5 patients (Stroke 2004;35:1647). These results suggest that more extensive cardiac monitoring is likely to identify atrial fibrillation in more patients than conventional monitoring. It would be helpful to limit such intensive recording to patients who are at increased risk to have paroxysmal AF.

In this background, Wallmann et al prospectively studied ischemic stroke patients without known AF. They had previously found that the presence of frequent atrial premature beats (APBs) was associated with an increased incidence of paroxysmal AF in acute ischemic stroke patients. In their new study, they stratified patients for the presence or absence of frequent APBs (> 70 per 24 hours). Patients were excluded from the study if atrial fibrillation was found on a 24-hour ECG recording or documented by other means during their admission. Patients were also excluded if they had severe aphasia, other severe cognitive deficits, or a life expectancy of less than 6 months. One hundred and twenty-seven patients were enrolled and underwent 7-day event-recorder monitoring at 0, 3, and 6 months after their stroke.

The chief finding was that paroxysmal AF was found in 26% of patients with frequent APBs (13/50) and in only 6.5% of the remaining patients (5/77). This finding was highly significant with P = 0.0021. Multivariate analysis, which looked at 12 other clinical variables, demonstrated that the presence of frequent APBs was the only independent predictor of paroxysmal AF (P = 0.01). Of the patients with frequent APBs, 8% were found to have AF on the first 7-day event-recorder, another 12% on the second one, and another 6% on the third one.


These results are important because the presence of atrial fibrillation in stroke patients is generally considered to be a strong indication for oral anticoagulation unless there is a clear contraindication. The results, therefore, suggest that more prolonged monitoring than is generally done may be appropriate, especially in patients with frequent APBs, although AF was even detected in a small percent of patients without frequent APBs.

As the authors correctly point out, however, the presence of AF does not necessarily mean that it is the cause of the patient's stroke, and the value of anticoagulation has not been tested in patients who have been found to have AF only on the type of long-term monitoring done in this study. As the authors suggest, a definitive answer to the question of optimal treatment for such patients would require new clinical trials. Nevertheless, in the absence of such trials, available data about the benefit of anticoagulation in patients with AF suggest that anticoagulation should be considered seriously even if AF is detected only with long-term monitoring.