Patent foramen ovale and infarct volume in cryptogenic stroke

Background: The causal relationship between patent foramen ovale (PFO) and stroke is controversial. We hypothesized that if PFO is a pathway of embolic source, there might be a correlation between PFO characteristics (ie, size or extent of shunt) and ischemic lesion burden (ie, infarct volume and number). Methods: From ischemic stroke patients admitted to Asan Medical Center between January 2000 and October 2007, we identified those who had (1) acute ischemic lesion on diffusion-weighted imaging within 5 days of symptom onset and (2) cryptogenic stroke and only PFO detected by transesophageal echocardiography. PFO characteristics on echocardiographic studies included size, shunt grade, shunt pattern, and the presence of atrial septal aneurysm (ASA). Results: Enrolled were 75 patients (male, 56%; mean age, 45.3 ± 13.9 years), including 10 patients (13.3%) with ASA. In univariable analysis, PFO size was positively correlated with log-transformed infarct volume (LIV) (regression coefficient =.469, P =.009). After adjusting for hypertension, stroke history, and migraine (all P <.2), PFO size remained independently associated with LIV (regression coefficient =.481, P =.007). Lesion number was negatively correlated with PFO size (Spearman coefficient rho = -.251, P =.03). The initial National Institutes of Health Stroke Scale scores tended to be positively correlated with PFO size (Spearman coefficient rho =.223, P =.054). Conclusions: In cryptogenic stroke, PFO size and ischemic lesion burden were positively correlated. These results support that PFO may play a role as a pathway of embolic source in cryptogenic stroke.