Do patients with cryptogenic stroke or transient ischemic attack have fewer recurrent strokes with patent foramen ovale closure or medical treatment?

In this limited systematic review, patients with cryptogenic stroke or transient ischemic attack (TIA) undergoing patent foramen ovale (PFO) closure have fewer subsequent strokes than those treated medically, at the risk of developing atrial fibrillation. This paper does not address whether outcomes vary by PFO size. (LOE = 1a-)

Ntaios G, Papavasileiou V, Sagris D, et al. Closure of patent foramen ovale versus medical therapy in patients with cryptogenic stroke or transient ischemic attack: updated systematic review and meta-analysis. Stroke 2018;49(2):412-418.  [PMID:29335335]

Meta-analysis (randomized controlled trials)

Unknown/not stated

Various (meta-analysis)

These authors searched PubMed for randomized trials that compared medical treatment with PFO closure in patients with cryptogenic stroke or TIA. They only included studies published in English. Other than looking at reference lists of relevant papers, they did not do much to identify unpublished studies (eg, clinical trials registries, gray literature). Although this journal does not routinely report funding sources, one of the authors reported ties to industry. The authors don't describe if the process of article selection was done in a masked, independent manner and don't describe assessing the risk of bias in the included trials. Ultimately they included 5 studies with 3627 patients who were followed up for a median 3.7 years (range 2 - 5.9 years). Slightly more than 3% of the patients had a subsequent stroke and those undergoing PFO closure had fewer strokes than those treated medically (2% vs 4%; number needed to treat = 47 for 3.7 years; 95% CI 31 - 98). The authors found no significant difference in all-cause mortality or in the rate of myocardial infarctions or TIAs. More patients undergoing PFO closure developed incident atrial fibrillation than those treated medically (1.3 vs 0.25 per 100 patient-years; number needed to treat to harm = 27 for 3.7 years; 21 - 37). The authors report that atrial fibrillation after PFO closure was transient, with nearly 75% resolving after 45 days, but they provide no similar data for patients treated medically. The studies demonstrated heterogeneity among the findings. Finally, the authors don't mention whether the studies addressed the size of PFO defect and outcomes.