I. Pathophysiology

  1. The foramen ovale runs between the left and right atria and normally closes soon after birth
  2. In 15-30% of humans it stays patent throughout life (mean diameter 4.9mm in one study), providing a potential shunt between the atria
  3. Found more often in pts with idiopathic ischemic cerebrovascular accident than in control subjects  (Circ. 105:2580, 2002--JW).  However, a causal relationship is unclear and PFO may not be a sig. predictor of recurrence of idiopathic CVA.
    1. However, in a prospective study of 601 pts with ischemic CVA (mean age 59yo), all of whom underwent transesophageal echocardiography, risk of recurrent CVA over 2y f/u were not more common in those with patent foramen ovale or atrial septal aneurysm on TEE (J. Am. Coll. Cardiol. 42:1077, 2003--JW)
    2. Also, in a case-control study comparing pts with cryptogenic vs. non-cryptogenic CVA and controls without history of cerebral ischemia (all undergoing transesophageal echocardiography), after adjustment for potential confounders, the prevalence of patent foramen ovale was not sig. higher in pts with cryptogenic CVA (  (Mayo Clin. Proc. 81:602, 2006--JW)
    3. In a meta-analysis of four controlled observational studies involving 1,081 pts with prior cryptogenic CVA or TIA, over mean 31-42 f/u, there was no sig. diff. among pts with or without PFOs in risk of recurrent CVA or TIA. (Neurol. 73:89, 2009-JW)
  4. Frequently associated with atrial septal aneurysm, itself associated with risk of CVA

II. Diagnosis

  1. Transesophageal echocardiography with "bubble study" (injection of microbubble contrast agents into systemic vein) is more sensitive than standard transthoracic echo
  2. Sensitivity increased with Valsalva and coughing during echo

III. Treatment

  1. Antiplatelet therapy for secondary prophylaxis against CVA for patients with PFO and a history of CVA
  2. Impact of PFO closure on risk of CVA
    1. 308 pts with idiopathic CVA and PFO who underwent either medical tx of PFO closure; over 4y f/u, the PFO recipients had no sig. diff in incidence of (death, CVA, or TIA), though they had larger mean right-to-left shunt and higher prevlance of having had > 1 CVA than the medical-treatment group (J. Am. Coll. Cardiol. 44:750, 2004-abst)