Residual shunting after device implantation is another significant complication that may result in recurrent thromboembolic events. Interatrial shunting may occur around or through the device.

As many of the patients who undergo percutaneous PFO closure do so for presumed paradoxical emboli and not for hemodynamic compromise, and since the PFO patients usually have a small shunt at baseline, it seems reasonable not to label any procedure with a residual shunt as a clinical success.

Residual shunts are mostly identified during routine postprocedure TEE with color Doppler echocardiography or using agitated saline contrast injection, at rest and with the Valsalva maneuver. Manifestation of new or recurrent symptoms should prompt a thorough evaluation of the closure device by TEE to assess for a new or worsening shunt.

Small residual right to left shunt (rRLS) is not uncommon after device implantation but large rRLS is rare (less than 3 % of RLS-positive patients after 1 year) [12]. rRLS sizing is assessed semi-quantitatively using saline-enhanced transcranial Doppler (TCD), by assuming a cut-off of more or less than 10 bubbles recorded in the cerebral vessels as a criterion to discriminate large versus small shunt [13].

Complete closure after single device implantation was more common with the Amplatzer and with the CardioSEAL-STARflex than with the Helex occluder: Amplatzer vs. Helex vs. CardioSEAL-STARflex: n = 217 (98.6 %) vs. n = 202 (91.8 %) vs. n = 213 (96.8 %; P = 0.0012) [9]. The higher rates of residual shunts with the Helex may be related to the design of the device and possibly to decrease rate of fibrosis by the Helex device [14].

The clinical implication of large rRLS is debatable; the preliminary results from Italian Patent Foramen Ovale Survey documented 14 cases of neurological recurrence at 1-year follow up, no large rRLS was reported amongst these cases [12]. Another study supported the aforementioned findings by studying the recurrence of embolic events after transcatheter closure of PFO in 1930 patients and concluded that there was no significant association between the presence of a residual shunt and recurrent events as the majority of events occurred in patients without residual shunts (54/63) [hazard ratio (HR) 1.7; 95 % confidence interval (CI) 0.8–3.6, P = 0.16] [15]. On the other hand, the amount of residual shunting was a predictor for recurrent embolic events or the reason behind volume overload of the right ventricle in other published studies [13, 16].

The optimal management of large rRLS should be individualized, some might benefit from implantation of an additional closure device. Diaz et al. evaluated 424 patients who underwent PFO closure between 1995 and 2007, 5 % of those patients had moderate to large residual shunt detected on transthoracic 2D echocardiography with administration of agitated saline contrast and color flow Doppler at 6 months follow up [17]. Of the 21 study patients with moderate-to-large residual shunt, 20 underwent transcatheter closure using a second device. The technique was successful in 95 % (19/20), and all but one patient had complete shunt closure at 6 months of their percutaneous reintervention. No recurrent stroke or transient ischemic attack (TIA) was observed over a follow up period of 2.9 years. The only independent predictor of residual shunting at 6 months in this study was the presence of 24-h postprocedure shunt. Another study identified the presence of atrial septal aneurysm and longitudinal Fossa Ovalis dimension >20.8 mm as independent predictors of significant residual RLS at 12 months (OR 7.6; 95 % CI: 1.38–42.35; p = 0.02 and OR 8.5; 95 % CI: 1.55–46.95; p = 0.014 respectively) [18].

Tachyarrhythmia, mainly atrial fibrillation and bradyarrhythmia, mainly atrio-ventricular block (AVB) are the two most common arrhythmic complications of percutaneous PFO closure.

Atrial fibrillation occurs at a rate of 2–5 % after device implantation [14, 19]. Most of the episodes of atrial fibrillation occur within the first 4 weeks and the majority within 6 months following the procedure (62.3 % and 15 % respectively) [19]. The incidence was more common with the STARFlex devices. Incidence was higher among elderly patients [19]. In one study, 1,349 consecutive patients underwent PFO closure with different devices, 53 patients developed new onset atrial fibrillation, of those 56.6 % developed chronic atrial fibrillation, whereas the rest had a single paroxysm lasting less than 48 h that terminated either spontaneously or with cardioversion [19].

Development of new-onset 1st degree (AVB) after the procedure was associated with an increased risk of atrial arrhythmia post-procedure as well [20].

The development of atrial arrhythmia following device implantation may be related to the fact that these devices cause local irritation of the adjacent myocardium and hence increase the automaticity or enhance the triggered activity in the surrounding area by focal irritation. Another proposed theory is that these devices act as an anatomical obstacle that provides the substrate for the development of anatomic macro-reentry [21].

The most concerning complication is complete AV block. Complete AVB has been described sporadically in several studies and case reports, in most of these cases the AVB was transient and recovered in a short period of time, however a few cases required permanent pacemaker implantation [20, 22].

Suda et al. [23] suggested that a larger shunt (Qp/Qs ratio 2.8 ± 0.9 vs. 2.1 ± 0.8, p < 0.01) and larger device size (24 ± 5 vs. 19 ± 6 mm, p < 0.01) were the only determinant factors for AVB in ASD patients.

As with other percutaneous cardiac interventions, there is a risk of pericardial tamponade during percutaneous PFO closure. It is estimated in the range of 0.5–1 % [26] and results most commonly from perforation of the left atrial appendage during fixing of the trans-septal guide wire [27]. Perforation of other structures can be responsible for this complication, though less commonly (right ventricle, right atrium, pulmonary veins) [28]. Tamponade may occur as a late complication, if the device erodes through adjacent cardiac tissue [27].

Cardiac erosion after PFO closure is rare. It has been described after ASD closure but the data are sparse with few reports of erosion after closure of PFO. It should be noted that erosion has been described after use of an Amplatzer ASD device in a PFO. Device erosion resulting in tamponade was diagnosed 6 years following device closure of a PFO with a 14-mm Amplatzer septal occluder (ASO) device. The patient presented with pleuritic chest pain and was found to have pericardial tamponade that recurred after percutaneous drainage. Upon mediastinal exploration, the left atrial disk of the occluder device was noted to have eroded through the dome of the left atrium, with approximately 4–5 mm of the disk visible from outside the left atrium [29]. Another case of very late device erosion was recently reported and occurred 8 years following percutaneous placement of a 36 mm ASO device. This case represents the longest period between ASO device implantation and clinical presentation following erosion [30].

Perforation of the mitral valve caused by wire fracture is another complication that has been reported with the use of the HELEX device [31].