Background

New-onset atrial fibrillation (AF) after percutaneous closure of patent foramen ovale (PFO) was described in several previous studies. Predictors of AF are incomprehensively examined. Strategies for detection and treatment of AF after the procedure remain unclear, also differences among the devices and their ischemic outcome after neurological events. This meta-analysis aimed to investigate the incidence of post-interventional AF according to device type and analyze AF management strategies.

Methods

A systematic search was conducted in Cochrane, MEDLINE and EMBASE. A meta-analysis was performed evaluating controlled trials. The primary outcome was the incidence of atrial fibrillation. Further parameters, such as incidence of AF sorted by devices, AF screening strategy, treatment and neurological events, were extracted. New AF was determined as early onset within one month after implantation and late thereafter. The studies were separated by screening strategy, high-quality with external loop recorder or Holter-Electrocardiogramm (ECG), opportunistic with ECG and Holter-ECG led by symptoms and low-quality with 12-lead ECG.

Results

The search led to seven randomized controlled trials and 17 cohort studies that were eligible for quantitative analysis. 9979 patients received the percutaneous PFO closure after cryptogenic stroke or transient ischemic attack (TIA) except for 117 patients with different indication (migraine, decompression sickness, diving disease, platypnea-orthodoxia syndrome). The meta-analysis of controlled trials showed an AF incidence of 5.1% in the interventional and 1.5% in the conservative arm, respectively (OR 3.17, 95% CI 1.46-6.86, p=0.03, I²=55%). Only 362 patients received a high-quality surveillance strategy and the overall AF incidence was higher compared to the low-quality group (3.3-15% vs. 0.2-4.3%). A difference in early and late AF onset was not found, though, limited by short follow-up duration of the high-quality studies of up to 12 months. Regarding the devices, in almost half of the cases (49.5%) the Amplatzer device was used, followed by Helex (15.4%), CardioSEAL (11.2%), Starflex (9.4%) and Premere (7.2%). CardioSEAL had a higher AF incidence with 3.5-7.0%% compared to the other most frequent used devices except for one study in the Amplatzer device (0.4-4.8%). 19 trials reported 307 patients with 313 AF episodes, where 172 (56%) had a documented AF treatment strategy. Besides spontaneous conversion in 44 patients (25.5%), 111 (64.5%) received a rhythm and 17 (10%) a rate-control strategy. Eleven of the 19 trials reported an oral anticoagulation in 72 patients (23.5%). 16 (22.2%) had a short-term therapy with less than 6 months. 56 (77.8%) got long term anticoagulation mostly due to additional cardiovascular risk factors. Overall, the cumulative incidence of TIA and stroke was 1.5%, and in the AF treatment cohort of 4278 patients 1%, showing no difference despite gaps in detection and treatment. Six patients with detected AF had a documented neurological event, one under oral anticoagulation in his second AF episode after implantation.

Conclusion

Patients after percutaneous PFO closure had a higher AF incidence compared to the conservative arm. Heterogeneity in surveillance and follow-up strategy limit the generalizability, hence a standardized follow-up is required for definite interpretation. The clinical benefit by reduction of strokes in AF patients after procedure remains unclear.