Patent foramen ovale in a large population of ischemic stroke patients: Diagnosis, age distribution, gender, and race

Background: Patent foramen ovale (PFO) is a well-recognized risk factor for ischemic strokes. The true prevalence of PFO among stroke patients is still under debate. Transesophageal echocardiography (TEE) is the "gold standard" in diagnosing PFO but the physiology requires right-to-left atrial shunting. In this report, we evaluate the prevalence of PFO in a diverse group of ischemic stroke patients studied by TEE. Methods: TEE of 1,663 ischemic stroke patients were reviewed for cardiac source of embolism, including PFO and atrial septal aneurysm (ASA). Agitated saline bubble injection was performed to look for right to left atrial shunting. Success of maneuvers to elevate right atrial pressure (RAP) was noted by looking at the atrial septal bulge. Results: Among 1,435 ischemic stroke patients analyzed, the presence or absence of PFO could not be determined in 32.1% because bulging of the septum could not be demonstrated in patients with negative contrast study despite aggressive maneuvers to elevate RAP. Of the remaining 974 patients, 294 patients (30.2%) had a PFO. The mean age was 61.5 years in both groups, with a bimodal distribution of PFO and the highest prevalence occurring in <= 30-year-old group. Prevalence of PFO was similar in men (32.4%) and women (28.15%, P = 0.15); and in Caucasian (32.1%) and African American (27.7%; P = 0.15). ASA was present in 2.02% and hypermobile septum in 2.49% of the 1,435 patients. PFO was seen in 79.3% of the patients with ASA. Conclusion: Successful elevation of RAP cannot be achieved in a significant number of patients undergoing TEE and determination of PFO may be difficult. In our series, the true prevalence of PFO among ischemic stroke patients was 30.2% taking into account only those patients who showed no shunting despite bulging of the atrium septum into the left atrium (PFO absent group) during the contrast study. There was no gender or racial difference in the prevalence of PFO, but there was a bimodal distribution in prevalence with age.