Context The role of percutaneous closure of patent foramen oval (PFO) in patients with cryptogenic stroke has been very controversial for years due to a lack of clear evidence.

Objective Systematic review and meta-analysis of the effect of percutaneous PFO closure for secondary prevention of cryptogenic strokes as compared to best medical therapy (BMT).

Data sources Trials were identified through a literature search until 28 May 2013.

Study selection Controlled clinical trials (randomised and non-randomised) comparing percutaneous PFO closure with BMT.

Data extraction and synthesis Main end point of interest was stroke. A random effects model was used to calculate the pooled relative risks (RR) with 95% CIs.

Results A total of 14 studies (three randomised controlled trials (RCT) and 11 non-randomised observational studies (non-RCT)), and a total of 4335 patients were included for this analysis. There was no significant treatment effect of PFO closure regarding stroke among the RCT (RR 0.66, 95% CI 0.37 to 1.19, p=0.171). However, among non-RCT stroke was reduced (RR 0.37, 95% CI 0.20 to 0.67, p<0.001) after PFO closure. A time-to-event (stroke) analysis, combining all three RCT and the two non-RCT which applied strict multivariate adjustments, showed a borderline significant risk reduction after PFO closure (HR 0.58, 95% CI 0.33 to 0.99, p=0.047). Neither risk of bleeding nor mortality differed significantly between the groups. However, there was a higher incidence of new onset atrial fibrillation in the closure group (RR 3.50, 95% CI 1.47 to 8.35, p=0.005).

Conclusions Percutaneous closure of PFO in patients with cryptogenic stroke does not appear superior to medical therapy according to currently available randomised data. Furthermore, it is associated with an increased incidence of atrial fibrillation. However, there are signals pointing towards a potential benefit and more research should be strongly encouraged.