The Closure of Patent Foramen Ovale in Preventing Subsequent Neurological Events: A Bayesian Network Meta-Analysis to Identify the Best Device.

BACKGROUND

Randomized-controlled trials (RCTs) reported a finding on the safety and efficacy of percutaneous patent foramen ovale (PFO) closure to prevent stroke recurrence. It showed that the Amplatzer (AMP) device appears to be superior to medical therapy (MT) in preventing strokes and episodes of atrial fibrillation (AF), than other devices. We performed a network meta-analysis (NMA) to evaluate the closure of PFO in preventing subsequent neurological events while investigating the results obtained by specific devices.

METHODS

We searched 3 databases (MEDLINE, EMBASE, CENTRAL/CCTR) and identified 6 RCTs until March 2019. We performed an NMA and used pooled ORs. Analyses were done in NetMetaXL1.6-WinBUGS1.4.

RESULTS

Six RCTs with 3,560 patients (mean age 45.2-46.2 years) were included in the present NMA. Depending on the device, 4 groups of patients were compared with MT: 1,889 patients undergoing PFO closure were significantly less likely to experience a stroke than 1,671 patients treated with MT (ORs 0.41; 95% Cr.I. 0.27-0.60 with fixed-effects model and ORs 0.22; 95% Cr.I. 0.05-0.70 with random-effects model). The patients with AMP showed a similar risk than those treated with Helex/Cardioform (HLX/CF) or with a group of 11 multiple devices. This suggests the equality between the 2 most currently used devices. When assessing TIA and, for the safety analysis, major bleeding, both models confirm no significant difference between any devices and MT. PFO closure increased the risk of new-onset AF: MT induces AF significantly less than all the devices. In favor of the AMP, there is a reduced number of cases of AF versus MT; however, no device superiority has been established in comparing HLX/CF and other devices in a random effect model.

CONCLUSIONS

Our NMA provides evidence in favor of PFO closure with all the devices currently in use. We can conclude that these devices are better than MT, but not that 1 device is better than the rest in reducing stroke recurrences and AF episodes in the follow-up.