T he frequency of patent foramen ovale (PFO) in the general population has been estimated to be 15% to 35%. 1 In patients with cryptogenic stroke, PFO is more common, but a clear causative relationship is not well established.2 Recurrent cerebrovascular events are frequent in medically treated patients with a history of paradoxical embolism 3 ; however, in more than one third of these patients concurrent pathogeneses (other than the sole presence of PFO) are identified. 4,5 Thus, it still remains unclear whether PFO is causally related to firstever or recurrent cerebral ischemia or constitutes an incidental finding of diagnostic work-up. 6,7 PFO size and the degree of functional shunting could represent potential risk factors for cerebral ischemia in medically treated patients with PFO.8 PFOs in patients with cerebral ischemia are thought to be larger 9 and to be more frequently associated with atrial septal aneurysms, compared with the PFOs in asymptomatic patients. 10 Moreover, in transcranial Doppler (TCD) studies, large PFOs were found to be significantly associated with more microembolic signals compared with small PFOs.11,12 However, data from other studies suggest that the shunt grade assessed with TCD is not associated with the brain infarct volume on computed tomography 13