Device Closure of Patent Foramen Ovale—Is it Time to Update the Guidelines?

ABSTRACT

Background: The current American Association of Neurology guidelines do not recommend patent foramen ovale (PFO) closure for prevention of cryptogenic strokes, however recent findings from randomized control trials (RCT) have challenged this recommendation. We performed a comprehensive analysis of RCTs available to date to compare PFO closure to medical therapy in patients with cryptogenic stroke.

Methods: Five databases were analyzed including EMBASE, PubMed, Cochrane, Scopus and Web-of-Science from January 2000 to September 2018. We included RCTs comparing device closure of PFO versus medical therapy in patients with cryptogenic stroke. Heterogeneity was determined using Cochrane’s Q statistics. A random effects model was used.

Results: Six randomized controlled trials and 3,560 patients were included in the analysis. Mean follow-up was 47 ± 19 months. Device closure was associated with a 42% reduction in recurrent embolic strokes when compared to medical therapy (risk ratio (RR): 0.58, 95% CI: 0.39–0.88, p = 0.009). This occurred despite atrial fibrillation occurring at a higher rate with device closure when compared to medical therapy (RR: 4.75, 95% CI: 2.35–9.60, p < 0.001). There was no difference between the two groups with respect to all-cause mortality, transient ischemic attacks, major bleeding or adverse events.

Conclusions: PFO device closure in appropriately selected patients with cryptogenic stroke is associated with a significant reduction in the risk of recurrent ischemic stroke. The increased incidence of atrial fibrillation associated with device closure appears to be transient. Long-term monitoring may be essential.

KEYWORDS:

• Patent foramen ovale
• cryptogenic stroke
• meta-analysis

Disclosure statement

Dr. Garcia is a consultant for Surmodics, Osprey medical, Medtronic, Edwards Lifesciences, Abbott and Boston Scientific. Research grants from Edwards Life Sciences, Minnesota Veterans Research Foundation and VA Office of Research and Development. Other authors have no relevant disclosures.

Supplemental material

Supplemental data for this article can be accessed here.