Cryptogenic Strokes

Myth or Evidence-Based Reality: Do Patent Foramen Ovales Lead to Cryptogenic Strokes?


Is routine closure of patent foramen ovale warranted?

On December 18, 2005, Ariel Sharon, Prime Minister of Israel, experienced the sudden onset of aphasia.1 Despite being overweight, he had none of the traditional risk factors for cerebrovascular disease—hypertension, history of smoking, diabetes, or elevated cholesterol levels. When further workup was pursued, transesophageal echocardiography revealed a large patent foramen ovale (PFO), with an atrial septal aneurysm and spontaneous right- to-left shunt.

It was strongly suspected that he had had a paradoxical embolus (right circulation to left through his PFO), followed by a dominant hemispheric stroke. He received anticoagulant therapy; unfortunately, before closure of the PFO could be accomplished, he had a debilitating right hemispheric hemorrhage. He remains in a vegetative state as a result.

Although Sharon’s clinical story supports a PFO as the source of his stroke (because of the additional findings of an atrial septal aneurysm and spontaneous right-to-left shunt), is there evidence demonstrating a relationship between venous thromboembolism (VTE) and strokes in general, especially as a result of PFOs without atrial septal aneurysms and spontaneous shunting? Since PFO closures are on the rise, the answer has become more than academic.

This month’s “Top Paper” retrospectively looked at 187,000 patients (from 1979 to 2006) who sustained a VTE complicated by an ischemic stroke.2 In this cohort, clinical suspicion would be raised for a right-sided clot traversing a PFO and secondarily causing the stroke. Surprisingly, in the overall cohort of 11,390,000 patients, ischemic strokes were more common in those without a VTE. If PFO was a significant cause of ischemic stroke, the data would be completely reversed. In fact, the group with VTEs had fewer strokes.

Looking at the numbers from the perspective of PFO prevalence is also beneficial. About 200,000 cryptogenic strokes occur each year in the United States. With the known prevalence of PFOs, that would suggest about 70,000 to 90,000 strokes per year result from PFOs and paradoxical emboli.2 The data do not bear out a cause-and-effect relationship.

If an association between PFOs and cryptogenic strokes in the setting of VTE cannot be proved, why should PFOs be closed percutaneously? Dalen and colleagues2 note that 4046 PFO closures were performed in the United States in 2004. These closures came at a price. There was a 6% to 10% complication rate (including transfusion, tamponade, and urgent surgery). From a financial perspective, the “price tag” was also impressive. Percutaneous, transvenous closure without complications costs approximately $28,466 per patient and $61,325 if complications ensue.

The authors concluded, “Twenty-four years after the first report of closure of PFOs to prevent recurrent stroke, there is still no evidence that patients who have a PFO are at increased risk of ischemic stroke due to paradoxical embolism.”2 Although Ariel Sharon’s history may be categorically different because of his additional aneurysm and spontaneous shunt, should there be a moratorium on PFO closures pending the results of randomized clinical trials already in progress? Maybe George Orwell was right when he said, “When myths are believed in, they tend to become true.”3


1. Ben-Hur T, Lotan C, Naparstek Y. Ariel Sharon’s stroke: the treatment he received—and why. Neurol Today. 2006;6:8.
2. Dalen JE, Stein PD, Matta F. Patent foramen ovales and cryptogenic strokes: another look. Am J Med. 2011;124:2-3.
3. Orwell G. The Collected Essays, Journalism and Letters of George Orwell. Vol 3. Orwell S, Angus I, eds. New York: Harcourt Brace Jovanovich; 1968:6.