Timing on patent foramen ovale closure: early versus delayed closure

Abstract

Abstract Introduction Pivotal randomized clinical trials showed benefit in patent foramen ovale closure (PFOC) up to 9 months after a cryptogenic stroke or transient ischemic attack (TIA) [1-4]. It might be sensible to indicate PFOC also in non-recent cryptogenic strokes, but evidence is lacking. Purpose We aimed to compare the effectiveness of PFOC according to the time elapsed from the patient’s last episode of cryptogenic cerebrovascular event (CVE) or systemic embolism (SE). Methods A retrospective, international, multicentre, cohort study was performed to compare the outcomes of an early closure (EC, <9 months) or delayed closure (DC, ≥9 months) of patent foramen ovale (PFO) as secondary prevention after the index event. The primary end-point was recurrence of CVE or SE after PFOC. Results A total of 325 and 171 patients were classified in the EC and DC groups, with a median time to PFOC of 3.4 (IQR 1.9-5.5) and 17.5 (IQR 12.2-33.4) months, respectively. No significant differences were observed regarding the type of event indicating PFOC (stroke was the most frequent (77.8% in EC group vs 79.0% in DC group), followed by TIA (19.4% vs 18.7%)); neither about neuroanatomic localization of the lesions, nor contrast transcranial Doppler patterns, nor Risk of Paradoxical Embolism (RoPE) score (7 (IQR 5-8) vs 6 (IQR 5-7), p=0.077). Echocardiographic PFO characteristics of both groups were similar, except for a larger defect size (tunnel width) in the DC group (6 (IQR 4-14) vs 12 (IQR 6-16) mm, p=0.005)). PFOC device, procedural success (99.4 in EC vs 98.8% in DC group) and periprocedural complications (2.1 vs 0.8%) were comparable in both groups. Median follow-up was 2.0 (IQR 1.3-3.8) years in the EC group and 2.2 (IQR 1.2-4.9) in the DC group (p=0.721). No significant differences were observed concerning recurrence of stroke, TIA or SE (3.9% vs 2.6%, p=0.443), time from PFOC to recurrence (0.9 (IQR 0-2.1) vs 6.1 (2.5-8.8) years, p=0.090), death (1.4% vs 1.0%, p=0.697), residual shunt 12 months after PFOC, or antithrombotic treatment during follow-up (Figure 1). Moreover, a subanalysis considering those patients with a very delayed PFOC (≥24 months) after the index event also showed no differences in recurrence rate during follow-up compared to the earlier procedures (4.2% in the <24-month group vs 3.4% ≥24-month group, p=0.770). Conclusions No differences were observed regarding recurrences of CVE/SE in patients undergoing PFOC before and after 9 months from the index event. Despite the absence of evidence from randomized trials, this study suggests that the preventive benefit of PFOC after a cryptogenic CVE/SE may be extended beyond 9 months after the last event.Figure 1:Incidence of recurre