A 45-year-old woman was referred to the cardiology service of the Guillermo Almenara Hospital for evaluation 3 months after presenting with a left parietal–cortical stroke (Figures 1A and 1B). On physical examination, the patient was alert and lucid, with decreased muscle strength in the right hemibody and preserved sensibility. The rest of the physical examination was normal. She had no cardiovascular risk factors (height 160 cm, weight 62 kg, BMI 24 kg/m2).
Investigations
The patient underwent transthoracic echocardiography (TTE) with an agitated saline test suggesting patent foramen ovale (PFO). The TTE was complemented by transesophageal echocardiography (TEE), which showed a PFO with high-risk anatomical features (i.e. passage of >20 bubbles/frame in the first 3–5 beats, atrial septal aneurysm [ASA], and tunnel length of 20 mm; Figures 1C and 1D). The patient also underwent carotid vascular ultrasound, which ruled out the presence of atherosclerotic plaques, and 24-hour Holter ECG monitoring three times, which demonstrated the absence of supraventricular arrhythmias.
Management
The patient had a Risk of Paradoxical Embolism (RoPE) score of 7 points (PFO-attributable fraction 72%) and a probable PFO-related stroke according to the PFO-associated Stroke Causal Likelihood (PASCAL) classification, so percutaneous closure of the PFO was decided on. Based on the characteristics of the PFO, closure was performed using a LifeTech CeraFlex PFO-3025 device, through a 12 Fr delivery sheath (Figures 2A and 2B). The device was prepared using cold saline (15°C), and the size of the release sheath was chosen based on the manufacturer’s recommendations. Prior to release, several attempts were made to accommodate the device, each preceded by complete flushing of the device out of the release sheath. Once the device was positioned, its stability was assessed using the Minnesota maneuver (gentle pulling and pushing on the delivery wire to ensure the device is firmly deployed and not likely to migrate).1
After device deployment, TEE demonstrated a concave deformation of the right atrial disc, which had not been evident during previous device positioning and was associated with a significant residual shunt (Figures 2C and 2D; Supplementary Video 1). Recapture of the same with snare was attempted without success and the failure was attributed to the deformation. Due to multiple failed implantations and prolonged radiation exposure, it was decided to suspend the procedure and for the case to be reviewed by the heart team.
After discussion of the case, expectant management with serial echocardiography was chosen, with the subsequent possibility of placement of a second device or surgical closure of the defect depending on the persistence of the residual shunt. At 1-month follow-up, echocardiographic assessment showed spontaneous resolution of the deformity, with adequate apposition and spatial morphology, and the absence of an interatrial shunt (negative agitated saline test for PFO; Figure 3 and Supplementary Videos 2 and 3).
The patient received dual antiplatelet therapy (aspirin plus clopidogrel) for 3 months, followed by clopidogrel only (75 mg once daily). Eight months after the procedure, the patient has no cardiovascular symptoms and has not experienced a recurrence of ischemic stroke.
Discussion
A number of anatomical features are associated with an increased risk of ischemic stroke associated with PFO, including ASA (defined by an excursion >10 mm), a moderate to severe shunt (the passage of >20 bubbles in 3–5 beats in an echocardiographic study or a curtain pattern or >10 high-intensity transient signals on transcranial Doppler), hypermotility of the interatrial septum, a large and long-tunneled PFO, the presence of a Eustachian valve, or a Chiari network.2–4 Some of these characteristics are also associated with greater complexity of the procedure and a greater risk of residual shunt.
Atrial septal occluder devices are made based on nitinol, the unique properties of which (shape memory, super-elasticity, and high biocompatibility) enable its use in medical applications, but are also associated with the risk of device deformations.5 The shape memory property is affected by temperature, such that nitinol deforms and remains misshapen at low temperatures but recovers its original form upon heating.5 The incidence of these deformations varies from 0% to 9.5% depending on the type of occluder.5 Some possible reasons for the loss of shape memory have been described, including disc restriction in atrial structures, the use of delivery sheaths larger than the recommended size (which may allow twisting of the device as it is advanced within the sheath), angulation or kinking of the delivery system, repeated release and retraction within the sheath, reduction of the ambient saline temperature in the cath lab, or a manufacturing defect.5
The most commonly affected disc is the left disc (48%), whereas the rate of isolated right disc involvement is 13%.5 The ‘cobra head’ deformity is the most frequent, constituting 90% of cases, but many other variants have been described (Figure 4).5–8 These deformities are associated with a prolonged procedure and an increased risk of residual shunt, with only 40% of these deformations being reversible with minimal manipulation or spontaneous resolution.5
There are a number of new techniques for the implantation of these devices that have been associated with successful closure of atrial septal defects associated with a lower risk of disc malalignment. One of these is the fast atrial sheath traction (FAST) technique, which consists of the rapid unsheathing of both retention discs, allowing simultaneous clamping of both sides of the atrial septum.9 Before rapid release, it should be verified that the tip of the sheath is in the center of the left atrium, in the direction of the left superior pulmonary vein, and that the center of the device is kept at the atrial septum just right to the mid-spinal line.9 This method may facilitate proper retention of disc formation and reduce the risk of deformation. Moreover, delivery systems play a key role in the success of the procedure. The original rigid delivery wires transfer considerable tension to the occluder and atrial septum, thus increasing the risk of embolization, deformation, and residual shunts, especially in patients with complex atrial defects. The new release systems, with increased flexibility at the wire tip, reduce tension, allowing a more controlled deployment sequence, aiding easier device positioning, and minimizing any unwanted traction on the implant.10 This will result in less stress on the device and a more favorable orientation with an identical position of the device after release.
Our patient had several high-risk anatomical features, which significantly influenced the outcome of the procedure. In our case, the cause of the deformation was multifactorial, including the long PFO tunnel, the large size of the device (which could have conditioned the entrapment of the left disc in the left atrial structures), the use of cold saline for device preparation, the prolonged deployment time, the multiple attempts to implant the device (which could have contributed to torsion of the device center, favoring its deformation), and the excessive traction that the delivery system applied on the device. Knowledge of all these factors, training in new device deployment techniques, and the use of novel and flexible release systems could have prevented the occurrence of this deformation.
Finally, although there is no clear explanation as to how the position of the device improved spontaneously, we believe it was due to prolonged exposure to the patient’s body temperature, which allowed the shape memory property to recover, and to the compliance of the PFO tunnel, which gave way and allowed the device to position properly.
Conclusion
We present the first report of a rare right atrial disc malformation, termed the ‘inverted tulip,’ occurring during percutaneous PFO occlusion.