Chirantan Banerjee, MD

Katsanos AH, Spence JD, Bogiatzi C, Parissis J, Giannopoulos S, Frogoudaki A, et al. Recurrent Stroke and Patent Foramen Ovale: A Systematic Review and Meta-Analysis. Stroke. 2014

PFOs have high prevalence in the general population, but have been found more often in patients with cryptogenic stroke, and with larger size and associated atrial septal aneurysm. One of the most controversial topics in the stroke community over several decades has been the issue of patent foramen ovale (PFO) and its association with cryptogenic stroke, as well as medical therapy v/s closure of the atrial septal defect.


In the last 3 years, 3 randomized controlled clinical trials have been published amid a lot of scrutiny and hype. CLOSURE I trial, published in 2012 was first of the three, and compared closure with StarFlex device to medical therapy in cryptogenic stroke patients with PFO, and found that stroke recurrence was low in this population, and there was no benefit of closure over medical therapy.  PC Trial and RESPECT trial, both came out last year, and compared PFO closure with the Amplatzer PFO Occluder (St. Jude Medical) to medical therapy in patients with cryptogenic stroke. There were no significant differences in the primary endpoint (composite of death, nonfatal stroke, TIA, or peripheral embolism) or the individual components of the endpoint in the PC trial. But lower than expected event rates were again seen. In the RESPECT trial, again there was no significant difference in the main, intention-to-treat analysis (composite of recurrent nonfatal ischemic stroke, fatal ischemic stroke, or early death after randomization), but prespecified per-protocol and as-treated analyses hinted at significant benefits for PFO closure. Also, secondary analyses suggested benefit in patients with substantial (grade 3) right-to-left shunt and atrial septal aneurysm.

 Thus, the results provided arguments to both advocates and skeptics. All 3 trials were plagued by slow enrollment rates (attributed to substantial off-label use), as well as low event rates. Advocates argue that the follow-up in the trials were not long enough to demonstrate benefit.  


In the current study, Katsanos et al. undertook a meta-analysis of all available prospective studies reporting recurrent stroke/TIA rates in cryptogenic stroke patient with PFO. They used PRISMA guidelines and identified prospective studies by literature search, and as well as obtained additional data from authors of the included studies. Out of potential 21 studies, they excluded studies without controls, those mixing endpoints of medically and surgically treated patients, and those with combined outcome. Appropriate statistical tools were used, including the mixed effects model to combine data, and Cochrane Q and I2 statistic for heterogeneity. Shunt size was dichotomized as small or moderate – large. 

They found that medically treated cryptogenic stroke patients with PFO did not have a higher risk of recurrent stroke (RR=0.85; 95% CI= 0.59-1.22; p=0.37) or recurrent stroke/TIA (RR=1.18; 95% CI, 0.78-1.79; p=0.43). Also, size of PFO was not associated with the risk of recurrent stroke/TIA. Despite the usual limitations of meta-analyses like selection bias, publication bias, attrition bias etc., these findings add significant weight to the growing argument against closure of PFO in cryptogenic stroke patients, which is also the current AHA guideline. 

Another meta-analysis by Spencer et al. published in the BMJ earlier this year had also concluded that there was insufficient support for closure as compared to medical therapy, even when the analysis is restricted to ‘per-protocol’ patients or patients with concomitant atrial septal aneurysm. The lack of association with PFO size in this study corroborates with findings from the ROPE database. 

Currently enrolling industry sponsored REDUCE randomized trial will evaluate closure by the Gore septal occluder v/s medical therapy. But, given the findings of the above studies, I am not sure it will tell us anything that we do not expect already!