Background: Percutaneous, mechanical closure of defects of the atrial septum fails to completely resolve shunting in up to 20% of cases. Little is known about the factors associated with device failure.
Methods: We measured the left atrial opening (X), right atrial opening (Z), tunnel length (Y), septum secundum, device-septum primum separation, and tunnel compressibility of the patent foramen ovale (PFO) in 301 patients with cryptogenic neurological events, PFO anatomy, and severe Valsalva shunting (Spencer Grade 5-5+). All patients then underwent percutaneous closure with the GORE®HELEX Septal Occluder device and were evaluated at 3 months for residual shunt by transcranial Doppler (TCD).
Results: Severe residual Valsalva shunt (TCD Grade 5-5+) was found at 3 months in 21 of 301 (7%) patients. X, Y, and Z were associated with failure with a high degree of statistical significance, whereas the width of the septum secundum, device-septum primum separation, and tunnel compressibility were not. An unanticipated finding was that 14 of 35 (40%) patients sized with a large balloon failed compared with 9 of 280 (3%) sized with a small balloon (P < 0.0001). In the multivariate logistic regression model, X (P = < 0.0001) and balloon size (P < 0.0001) were both strong predictors of failure.
Conclusions: In an intracardiac echocardiography-defined PFO population, characterized by severe baseline Valsalva shunt and a high incidence of persistent (rest) shunting, association of six intracardiac measurements to closure device failure by multivariate logistic regression showed that the width of the left atrial opening was a strong predictor of residual shunting. An unanticipated finding was that use of a large sizing balloon was also strongly associated with failure.
©2012, Wiley Periodicals, Inc.