By Edward P. Gerstenfeld, MD

SOURCE: Poddar KL, et al. Risk of cerebrovascular events in patients with patent foramen ovale and intracardiac devices. JACC Cardiovasc Interv 2014;7:1221-1226.

After some case reports of stroke due to electrophysiology (EP) device thrombosis in patients with a patent foramen ovale (PFO), concern has been raised about the risk of stroke with intracardiac devices in patients with known PFO. Thus, these investigators from the Cleveland Clinic did a retrospective database study of 2921 echocardiography-detected PFO patients and categorized them as having an EP device (231) or not. Patients with a stroke prior to detection of the PFO, or who had closure of the PFO during the follow-up period were excluded. The primary endpoint was ischemic stroke. Propensity scoring was used to match device patients to those without devices, which resulted in two groups with 231 patients each for the comparison.

Ischemic stroke occurred in 2.6% (n = 6) in both propensity-matched groups over an 11-year total follow up (minimum 4 years). A subgroup analysis of atrial fibrillation patients revealed a 7% stroke risk in both groups. The authors concluded that the risk of stroke among patients with an intracardiac EP device is the same whether they have a PFO or not.


As the use of EP devices has increased, concern has risen about the risk of stroke in patients with PFO who develop device lead thrombosis. This study allays those fears since the PFO-plus device group had the same stroke risk as those with a PFO alone. This is a retrospective database study that suffers from selection biases and unadjusted covariates. However, unlike other such studies, the investigators did propensity matching to minimize selection biases. Factors such as medication use (statins, oral anti-coagulants) and atrial fibrillation were matched between the groups. Also, patients with prior stroke were excluded because they have a higher incidence of another stroke. In addition, patients who had their PFO closed during the follow-up period were excluded. The major limitation of the study was the low event rate, but this, in itself, is reassuring.

The fact that atrial fibrillation was present in almost all of the stroke patients reminds us that there are several other causes of stroke in device patients that are more important than PFOs, whose relationship to stroke is controversial. Others include proximal aortic atheroma and left ventricular dysfunction or aneurysms.

The results are a relief since closing PFOs before placing an EP device or anticoagulating everyone with a device and a PFO would be challenging and of unknown value given the lack of data. Our efforts are better spent determining if other treatable conditions that raise the risk of stroke, such as atrial fibrillation, are present.