Patent Foramen Ovale and Cryptogenic Strokes in the Stroke in Young Fabry Patients Study

Background and Purpose— A patent foramen ovale (PFO) is disproportionately prevalent in patients with cryptogenic stroke. Without alternative explanations, it is frequently considered to be causative. A detailed stratification of these patients may improve the identification of incidental PFO. Methods— We investigated the PFO prevalence in 3497 transient ischemic attack and ischemic stroke patients aged 18 to 55 years in the prospective multicenter SIFAP1 study (Stroke in Young Fabry Patients 1) using the ASCO classification. Patients without an obvious cause for transient ischemic attack/stroke (ASCO 0) were divided into subgroups with and without vascular risk factors (ASCO 0+ and 0−). In addition, we looked for PFO-related magnetic resonance imaging lesion patterns. Results— PFO was identified in 25% of patients. Twenty percent of patients with a definite or probable cause of transient ischemic attack/stroke (≥1 grade 1 or 2 ASCO criterion; n=1769) had a PFO compared with 29% of cryptogenic stroke patients (ASCO 0 and 3; n=1728; P<0,001); subdivision of cryptogenic strokes revealed a PFO in 24% of 978 ASCO 3 patients (n.s. versus ASCO 1 and 2) and a higher prevalence of 36% in 750 ASCO 0 cases (P<0.001 versus ASCO 3 and versus ASCO 1 and 2). PFO was more commonly observed in ASCO 0− (n=271) than in ASCO 0+ patients (n=479; 48 versus 29%; P<0.001). There was no PFO-associated magnetic resonance imaging lesion pattern. Conclusions— Cryptogenic stroke patients demonstrate a heterogeneous PFO prevalence. Even in case of less conclusive diseases like nonstenotic arteriosclerosis, patients should preferentially be considered to have a non-PFO-mediated stroke. Clinical Trial Registration— URL: http://www.clinicaltrials.gov. Unique identifier: NCT00414583.