Epidemiology of Patent Foramen Ovale in General Population and in Stroke Patients: A Narrative Review.

Percutaneous closure of patent foramen ovale (PFO) in selected patients with cryptogenic cerebrovascular ischemic events (CEs) decreases the risk of recurrent stroke; however, optimal patient selection criteria are still under investigation. Candidates for PFO closure are usually selected from the pool of CE patients with a high risk of Paradoxical Embolism (RoPE) score. The RoPE score calculates the probability that PFO is causally related to stroke, based on PFO prevalence in patients with CE compared with that in healthy subjects. The latter has been set at 25% based on the average of autopsy and transesophageal echocardiography (TEE) studies. We conducted a comprehensive review of studies investigating PFO prevalence in general population and in patients with CE and non-CE using autopsy, TEE, transcranial Doppler (TCD) or transthoracic echocardiography (TTE). Studies were excluded if they (1) reported data from referred subjects with underlying cerebrovascular disease or (2) did not specify etiologically the events. In healthy/control subjects, PFO prevalence was 24.2% (1,872/7,747) in autopsy studies, 23.7% (325/1,369) in TEE, 31.3% (111/355) in TCD, and 14.7% (186/1,267) in TTE studies. All diagnostic modalities included PFO prevalence was higher in CE compared with healthy/control population [odds ratio (OR) = 3.1, 95% confidence interval (CI) = 2.5-3.8] and compared with non-CE (OR = 2.3, 95% CI = 2.0-2.6). In patients with CE, PFO prevalence in the young compared to the old was higher when the diagnostic modality was TEE (48.9 vs. 27.3%, < 0.0001, OR = 2.6 with 95% CI = 2.0-3.3) or TCD (58.1 vs. 41%, OR = 1.9, 95% CI = 1.6-2.5), but not TTE (53.3 vs. 37.5%, = 0.16). Regarding non-CE, PFO prevalence in the young compared to the old was higher when the diagnostic modality was TEE (20 vs. 12.9%, OR = 1.7, 95% CI = 1.0-2.8) but not TTE (10.4 vs. 7.8%, = 0.75) or TCD (22.8 vs. 20.1%, = 0.56). Given the limitations of autopsy and TEE studies, there is good reason not to take a fixed 25% PFO prevalence for granted. The estimation of degree of causality may be underestimated or overestimated in populations with PFO prevalence significantly lower or higher than the established. Given the high sensitivity, non-invasive nature, low cost, and repeatability of TCD, future large-scale TCD-based studies should investigate potential heterogeneity in PFO prevalence in different healthy racial/ethnic populations.