Which characteristics of patent foramen ovale and associated anatomical structures determine risk of cryptogenic brain ischemia?

Abstract Background Linking of the characteristics of persistent foramen ovale (PFO) and associated septal/atrial structures to a causal role in cryptogenic stroke is not definitely known. Purpose In a prospective study to assess the risk of morphological and functional features of PFO and selected septal/atrial structures for cryptogenic stroke or transient ischemic attack (TIA). Methods A total of 1 270 consecutive patients underwent contrast transesophageal echocardiography – 702 patients with cryptogenic brain ischemia and 568 controls without the history of stroke or TIA. We compared 224 patients with cryptogenic stroke or TIA and PFO (31.9% of all patients with brain ischemia) to 106 controls with PFO (18.7% of the control group) (p<0.001). We analyzed diameter and length of the PFO, presence of the large PFO (diameter ≥4 mm), long PFO (length ≥10 mm), atrial septal aneurysm (excursion ≥15 mm), Eustachian valve or Chiari network, shunt severity (graded with a 1–3 scale) and the right-to-left shunt present under basal conditions. Results All parameters of PFO significantly increased relatíve risk (RR) of cryptogenic stroke/TIA vs population without PFO [RR (95% CI); p]: PFO – 2.04 (1.57; 2.66); <0.001, large PFO – 4.99 (2.67; 9.39); <0.001, PFO with diameter ≥ median 2.4 mm – 3.20 (2.19; 4.68); <0.001, long PFO – 2.25 (1.65; 3.06); <0.001, PFO with length ≥ median 12 mm – 2.26 (1.61; 3.17); <0.001, moderate or large shunt – 2.25 (1.65; 3.08); <0.001, shunt present under basal conditions – 2.47 (1.74; 3.52); <0.001. None of the associated structures (atrial septal aneurysm, atrial septal aneurysm + PFO, Eustachian valve or Chiari network, atrial septal defect) significantly increased RR of cryptogenic stroke/TIA vs population without the respective structure. In a multivariable logistic regression model PFO diameter ≥ median 2.4 mm and long PFO emerged as significant risk factors of stroke or TIA. Patients with brain ischemia had significantly larger PFO diameter and higher prevalence of large PFO compared to controls (3.1±2.0 mm vs 2.3±1.4 mm, p<0.001; 8.8% vs 2.1%, p<0.001). Symptomatic patients had significantly higher prevalence of long tunnel (23.2% vs 12.3%, p<0.001) and of moderate/large shunt (22.2% vs 11.8%, p<0.001). There was no significant difference between patients and controls in prevalence of atrial septal aneurysm, atrial septal aneurysm + PFO, Eustachian valve or Chiari network, atrial septal defect. Conclusions Among the parameters of PFO large diameter was the most powerful risk factor of cryptogenic brain ischemia. Long PFO, moderate/large shunt and shunt present under basal conditions also significantly increased the risk. Associated septal/atrial structures were not found as risk factors. Funding Acknowledgement Type of funding source: None