Clinical outcomes and determinants of adverse events following percutaneous PFO closure: experience from a 227-patient cohort

Background

Percutaneous patent foramen ovale (PFO) closure is widely used to prevent recurrent cryptogenic stroke, yet real-world data describing patient characteristics and predictors of adverse outcomes remain limited. This study characterizes the clinical profile, procedural outcomes and follow-up events in a consecutive cohort undergoing PFO closure.

Methods

This single-center retrospective study included consecutive patients who underwent PFO closure between January 1, 2020 and October 31, 2025. Demographic, anatomical and clinical data were collected from electronic medical records and compared using appropriate statistical tests.

Results

A total of 227 patients were included (51% women; mean age 51 ± 13 years). The most common indication was ischemic stroke (82%). Comorbidities were relatively infrequent: dyslipidemia (35%), hypertension (27%), autoimmune disease (9%) and thrombophilia (7%). Atrial septal aneurysm was present in 50%, and a large right-to-left shunt in 68%.

Most procedures were perfomed using the Amplatzer PFO Occluder® (54%). Acute complications were rare (n=5): thrombus adherent to NobleStitch® (among 79 implantations), iatrogenic interatrial communication, takotsubo syndrome and two puncture-site hematomas.

Mean follow-up was 400 days. Only one case each of atrial fibrillation, stroke and minimal pericardial effusion occurred. Significant residual shunt requiring reintervention was observed in 7%; these patients were younger (41 vs. 51 years, p=0.008) and more frequently treated with NobleStitch® (92.9% vs. 33.5%, p=0.003).

Five patients were hospitalized during follow-up, all for non-cardiovascular causes. Hospitalized patients more frequently underwent urgent procedures (40% vs. 4.5%, p=0.006), were more often smokers (60% vs. 15%, p=0.031) and had a significantly longer PFO tunnel (p=0.033).

There were five deaths, all non-cardiovascular. Mortality was associated with older age (73 vs. 50 years, p=0.004), urgent procedures (p<0.001), aortic root dilatation (p=0.009), higher comorbidity burden (CHA₂DS₂-VA ≥3: 75% vs. 10%, p<0.001), absence of statin therapy (p=0.041) and platypnea–orthodeoxia as the indication (p<0.001).

Conclusions

In this real-world cohort, PFO closure was safe and effective, with very low rates of arrhythmia, stroke and periprocedural complications. Significant residual shunt was the only outcome directly related to procedural features, particularly younger age and NobleStitch® use. In contrast, hospitalizations and deaths—both non-cardiovascular—appeared to reflect overall clinical vulnerability rather than PFO-related anatomy. These findings reinforce the excellent safety profile of PFO closure and highlight the importance of individualized patient selection and follow-up.