How Silent Atrial Fibrillation Leads to Stroke

Stroke vs. Cryoptogenic Stroke

Morsa Images/Getty Images

In up to 40 percent of people who suffer a stroke, a definitive cause of the stroke cannot be identified even after a full evaluation. When a stroke occurs without a known cause, it is called a cryptogenic stroke.

The potential causes of cryoptogenic strokes, for the most part, are the same as for strokes in which the cause is successfully identified. Most cryptogenic strokes are probably due either to the formation of a thrombus (blood clot) in one of the brain’s arteries, or to an embolus (a blood clot that breaks loose from its site of origin and lodges in an artery somewhere else).

Distinguishing between these two kinds of strokes is important because the treatment to prevent a recurrent stroke is different. (Antiplatelet therapy with aspirin or clopidogrel is typically given after a thrombotic stroke, while anticoagulation with warfarin or one of the newer anticoagulant drugs is used after an embolic stroke.) Because the risk of recurrent stroke following a cryptogenic stroke is substantial—about 20 percent—it would be very helpful to know the underlying cause of the stroke so that steps could be taken to prevent another one.

Atrial fibrillation is a widely-recognized cause of embolic stroke. In atrial fibrillation, blood clots tend to form in the left atrium. If those blood clots break loose and travel to the brain, a stroke commonly occurs. Depending on other risk factors, the chances of a person with atrial fibrillation having a stroke can be as high as 18 percent per year.

This risk can be substantially reduced if an anticogulation drug is taken.

In many cases, atrial fibrillation is not chronic, but rather occurs only sporadically, and often infrequently and for relatively brief periods of time.

(This is called paroxysmal atrial fibrillation.) Furthermore, a surprising proportion of patients who have paroxysmal atrial fibrillation report no symptoms whatsoever from their arrhythmia. These patients are said to have occult, or silent, atrial fibrillation.

Silent atrial fibrillation has always been difficult to diagnose because it produces no symptoms and it is relatively unlikely to show up on an ECG, or even on a 24-hour Holter monitor study. However, in recent years several varieties of long-term ambulatory monitoring systems have been developed, that allow for ECGs to be continuously recorded for up to 30 days. Or, in the case of an implantable monitoring system (Reveal, by Medtronic), the ECG can be monitored for longer than a year.

Recently, two important studies were published which applied such long-term ambulatory monitoring systems in patients who had suffered cryptogenic strokes. Both studies found a surprisingly high incidence of silent atrial fibrillation in patients whose standard cardiac evaluations had been normal.

In the EMBRACE trial, 572 patients with cryptogenic stroke were randomly assigned to receive either 30-day ambulatory monitoring or a standard 24-hour Holter study.

Atrial fibrillation was detected in 16 percent of the patients having 30-day monitoring studies, but in only in three percent of the patients having Holter studies.

In the CRYSTAL AF trial, 441 patients with cryptogenic stroke, and whose 24-hour Holter monitoring studies showed nothing, were randomized to either no more routine monitoring or to the implantable Reveal monitoring device. After six months, atrial fibrillation had been detected in about nine percent of patients in the Reveal group, compared to only about one percent in the control group. And after a year, over 12 percent of patients in the Reveal group were discovered to have atrial fibrillation.

Therefore, it now appears that a substantial proportion of patients with cryptogenic stroke have silent atrial fibrillation. It is very likely (though not absolutely proven) that their strokes were embolic strokes caused by the atrial fibrillation, and that anticoagulation would reduce their risk of subsequent strokes.

The Bottom Line

Silent atrial fibrillation is far more common among patients who have had cryptogenic stroke than has been previously realized. Randomized trials are being designed to prove whether or not anticoagulation will significantly reduce the risk of recurrent stroke in these patients. However, those studies will require enrolling very large numbers of patients with cryptogenic strokes, determining which ones have silent atrial fibrillation, randomizing those who do have silent atrial fibrillation to therapy vs. no therapy, and then following them for a long period of time and measuring any reduction in subsequent strokes. This will be a major and very lengthy undertaking, so a definitive answer will not be forthcoming anytime soon.

Waiting for definitive studies may not be practical for somebody who already has had a cryptogenic stroke. Because stroke can be such a devastating condition, because recurrent strokes are so common following a cryptogenic stroke, because atrial fibrillation is so often the cause of stroke, and because the risk of stroke with atrial fibrillation can be substantially reduced with treatment, many experts would consider it entirely reasonable, even today, to use prolonged ambulatory monitoring to look for silent atrial fibrillation in anyone who has suffered a cryptogenic stroke.


Sanna T, Diener HC, Passman RS, et al. Cryptogenic stroke and underlying atrial fibrillation. N Engl J Med 2014; 370:2478.

Gladstone DJ, Spring M, Dorian P, et al. Atrial fibrillation in patients with cryptogenic stroke. N Engl J Med 2014; 370:2467.

Continue Reading