Atrial Septal Defect Occluder Device Embolization: Experience of a Tertiary Care Cardiac Center

ABSTRACT

Background:

Percutaneous device closure of atrial septal defect (ASD) has become an increasingly popular procedure as it offers several advantages. However, it is associated with infrequent, but life-threatening complications such as device embolization.

Objective:

To analyze the risk factors, common sites of embolization, associated complications, timing of embolization, and the treatment executed.

Settings and Design:

A retrospective study was performed at a tertiary referral center for cardiac services.

Material and Methods:

Pre-procedure, intra-procedure, and post-procedure data of patients whose ASD device embolized was collected retrospectively and analyzed for risk factors, common sites of embolization, associated complications, timing of embolization, and the treatment executed.

Results:

Thirty devices were embolized, out of which 13 were retrieved percutaneously in the Catheter laboratory, whereas 17 patients underwent surgery. Fourteen patients had an unfavorable septal morphology for device closure. Ten devices were embolized in the catheter laboratory, five in the intensive care unit, and two in the ward. The devices were embolized to almost all chambers of the heart and great vessels. One patient had an inferior vena cava rim tear while attempting percutaneous retrieval. One patient required a short period of total circulatory arrest (TCA) for retrieval of the device from ascending aorta, while another required a lateral position for retrieval from descending aorta. One patient required re-exploration for bleeding, while another had an air embolism and succumbed.

Conclusions:

Once embolization occurs, the risks associated increase manifold. Most of the surgical extractions are uneventful; however, there could be certain complications that may need repair of valvular apparatus, the institution of TCA, or the need for the lateral position. Air embolization though very rare can occur which could be fatal.

INTRODUCTION

Atrial septal defect (ASD) is a common congenital heart lesion with an incidence of 100 per 100,000 live births.[1] Surgical repair was the method of choice, until King et al.[2] performed the first transcatheter closure of an ASD. Percutaneous device closure of ASD has become an attractive option as it offers advantages of avoidance of surgery and cardiopulmonary bypass (CPB), short procedure time, short hospital stays, and lower rate of complications.[3]

Percutaneous closure of ASD offers several advantages; however, it is not without complications, the most dreaded of which is atrial septal defect occluder (ASO) device embolization. It was seen in 1.1% of the procedures in a published multi-center study, 0.2% of whom required surgical intervention.[3] The device may embolize anywhere such as to the right atrium (RA), right ventricle (RV), pulmonary artery (PA) or its branches, left atrium (LA), left ventricle (LV), and aorta. Embolization to certain sites such as PA,[4] left ventricular outflow tract and aorta can be life threatening. Percutaneous retrieval of the device may be attempted, but surgical retrieval is the mainstay of treatment.

We performed a retrospective study to analyze the risk factors, common sites of embolization, associated complications, timing of embolization, and the treatment executed for the embolized devices in a tertiary referral center for cardiac services.

MATERIAL METHODS

After obtaining Institutional Ethics Committee approval (SCT/IEC/1154/DEC 2017), we retrospectively collected and analyzed data from patients during the period from Jan 2005 to Oct 2017, in whom the ASD device was embolized. We collected demographic data, pre-procedure echocardiography data including the number and sizes of ASD, adequacy of rims, ventricular function, and right ventricular systolic pressure (RVSP). Intra-procedure data consisted of the type of anesthesia provided, monitoring modalities used, make/type of device deployed, timing and site of device embolism. Data was also collected after embolization of the device such as migration of the device, method of retrieval, and complications.

The techniques of anesthesia and deployment of ASO were as follows. As a protocol, transesophageal echocardiography (TEE) guidance was the first choice for ASO device closures at our institute, to minimize radiation exposure. Therefore, all the patients were intubated under general anesthesia and the TEE probe was inserted.

Procedure for device deployment

Through the femoral route, the defect (ASD) was crossed with a catheter and its tip was usually positioned in the left superior pulmonary vein (LSPV). A stiff guide wire was then placed in the LSPV and the delivery sheath passed over the stiff wire into the mouth of the LSPV. The dilator was removed from the sheath and the back bleed was allowed by placing the sheath below the level of the heart in a saline bowl, to prevent inadvertent sucking of air into the sheath resulting in air embolism. The ASO was passed through the delivery sheath and once positioned, the LA disc of the ASO was extruded into the LA and the LA disc was pulled back against the inter-atrial septum. The delivery sheath was peeled back over the loading cable to allow the release of the waist and then the RA disc. TEE was used to confirm proper placement and adequate capture of all margins and then the device was released from the loading cable.

When the procedure was uneventful, the trachea was extubated at the end of the procedure, a transthoracic echocardiogram (TTE) was performed to rule out ASO dislodgement, and then the patient was shifted to the cardiac intensive care unit (ICU) for overnight observation. The data was collected only using the hospital id no. into a hard disk by the principal investigator and stored for 3 years.

RESULTS

A total of 2237 patients underwent ASD device closure during the study period. Of them 30 (1.3%) ASD devices were embolized, out of which 13 (43%) were retrieved percutaneously in the Catheter laboratory, whereas 17 (57%) patients underwent surgery for extraction of the device. We further analyzed the data of 17 patients who came for the surgery. Only patients who are not suitable for percutaneous closure are taken up for primary surgical closure at our institute. A total of 541 patients underwent primary surgical closure of their ASD during the same period.

Pre-procedure data

The average age of patients was 23 ± 9 years (range 6–39 years). Ten of them were male whereas 7 were female. All of them were diagnosed with an ostium secundum ASD of size ranging from 6–30 mm (average 22.5 ± 6.3). Fifteen patients had a single ASD while the remaining two of them had two ASDs which were planned to be closed by a single ASO device. Pre-procedure TTE data revealed that most of the patients had less than ideal septum characteristics for device closure. Nine patients had deficient rims [Table 1], two had a floppy inter-atrial septum, whereas the septum was aneurysmal in three patients. All patients had good biventricular function with near normal RVSP (21–38 mm Hg).

Table 1.

Pre-procedure TTE data of the inter-atrial septum

Atrial septum characteristics	Number of patients
IVC rim deficient	5
Posterior rim deficient	3
Aortic rim deficient	1
Floppy inter-atrial septum	2
Aneurysmal inter-atrial septum	3

IVC=Inferior vena cava

Intra-procedure data

As a protocol, TEE guidance was the first choice for ASO device closures at our institute, to minimize radiation exposure. Therefore, all the patients were intubated under general anesthesia and TEE probe was inserted. When the procedure was uneventful, the trachea was extubated at the end of the procedure, the TTE was performed to rule out ASO dislodgement, and then patient was shifted to the cardiac ICU for overnight observation. In ten patients, the device had embolized while the patient was in the catheterization laboratory, and in seven patients, the embolization was detected after tracheal extubation when the patient was transferred to the cardiac ICU [Table 2]. Among the devices that were embolized in the catheter laboratory, one was embolized during deployment, seven soon after deployment, and two after extubation. While under anesthesia, all the patients had insignificant changes in vital parameters when the device was embolized, whereas device embolization in the post-extubation period was marked by the development of severe tachycardia in a patient and profound coughing in another. The TTE examination in both patients revealed embolization of ASO device. Among the seven patients, in whom the device embolization occurred after they were transferred out from the catheter laboratory, varying clinical presentations were observed. One of them developed desaturation to 90% about an hour after the transfer. Frequent ventricular ectopic beats, 5 hr after the procedure were noted occasionally in a patient, and frequently in another one. The other two complained of chest discomfort at 9 and 15 hr, respectively, after the procedure, whereas two patients remained asymptomatic and the device embolization was detected on echocardiography before hospital discharge approximately 15–48 hr after the device deployment. The make of the 17 devices that embolized were different, with six being Blockaid (Shanghai shape memory, Alloy Company Ltd., China), five Cocoon (Vascular Innovations Co. Nonthaburi, Thailand), one AmplatzerTM (Abbott Medical, Plymouth, MN, USA), while data of the remaining five could not be retrieved.

Table 2.

Timing and presentation of device embolization

Location	Timing	Presentation	Number of patients
Cath Lab	During deployment	Asymptomatic	1
	Soon after deployment	Asymptomatic	7
	Immediately after extubation	Tachycardia	1
	Immediately after extubation	Cough	1
Cardiac ICU	1 hr after procedure	Desaturation	1
	5 hr after procedure	Frequent VPCs	1
	5 hr after procedure	Occasional VPCs	1
	9 hr after procedure	Chest discomfort	1
	15 hr after procedure	Chest discomfort	1
Ward	16 hr after procedure	Asymptomatic	1
	48 hr after procedure	Asymptomatic	1

ICU=Intensive care unit, VPC=Ventricular premature contractions

The ASO devices had embolized almost all chambers of the heart and great vessels [Table 3]. Retrieval of embolized devices was unsuccessfully attempted in six patients in the catheter laboratory. One of them had an inferior vena cava (IVC) rim tear during the process.

Table 3.

Sites of Device embolism

Site	Number of patients
RA	1
RV	3
RA—later migrated to PA	1
MPA	4
Junction of MPA and LPA	1
LV	5
Proximal arch of aorta	1
Descending thoracic aorta	1

MPA=Main pulmonary artery, LPA=Left pulmonary artery

In the operation theater (OT)

Nine of the seventeen patients arrived in the OT with endotracheal tube in situ, whereas the remaining patients underwent an uneventful anesthesia induction and endotracheal intubation in the OT. All the patients received invasive arterial and central venous cannulas and were monitored using TEE. The pre-CPB period was uneventful in all except one who had episodes of ventricular tachycardia and required an emergency institution of CPB. In one patient the ASO device which was found in the RA in the catheter laboratory had migrated to the PA, while in the rest of the patients, no significant change in device position was noted. TEE also revealed a small thrombus on the ASO device in one patient. In two patients the device was abutting the AML and partially obstructing flow through the mitral valve. All patients underwent sternotomy, ASO device retrieval, and ASD closure except one who underwent a thoracotomy and device retrieval alone as the device had embolized to the descending thoracic aorta. One patient required a short period of total circulatory arrest (TCA) for the retrieval of the device from ascending aorta through an incision in close proximity to the aortic arch. The average aortic cross-clamp time was 23 ± 17 min and the CPB time was 49 ± 7 min. In one patient, the endothelium of the atrial septum was damaged and a small thrombus was found on the left atrial (LA) floor and LA disc of the device. Although the device was abutting on the mitral valve in two and the tricuspid valve in one patient, the valvular apparatus was intact in all and none required any valvular repair.

Perioperative complications

One patient required re-exploration for post-operative bleeding. Another patient while shifting on to the transportation trolley had a sudden episode of desaturation, hypotension, and cardiac arrest. TTE revealed air in the RA. She was resuscitated and transferred to the OT. During the pre-CPB period, her hemodynamic condition remained unstable and had an episode of ventricular tachycardia, for which her heart was defibrillated and emergency CPB instituted. The device was retrieved and the ASD closed uneventfully. However, she was the only patient, who required inotropic support of Dobutamine at 5 mcg/kg/min and Norepinephrine at 0.05 mcg/kg/min in the post-CPB period. The patient did not wake up in the post-operative period and a brain CT scan revealed a massive right hemisphere ischemic stroke with brain stem involvement. Subsequently, an electroencephalogram showed no electrical activity and the patient died after a few days.

DISCUSSION

Percutaneous ASD closure with an ASO is relatively safe and carries several advantages such as avoidance of prolonged anesthesia, shorter length of hospital stay,[3] avoidance of a surgical incision, and exposure to the CPB circuit with the associated adverse effects. The complications of percutaneous ASD closure include air embolism, vascular trauma resulting from large sheathes, device embolization, clot embolization, into the systemic circulation, occlusion of pulmonary or systemic venous return, perforation of the atrial septum, aortic perforation, infective endocarditis, atrial arrhythmia, device malposition necessitating retrieval, and delayed breakdown of device.[5]

ASO device embolization is a recognized complication of percutaneous transcatheter ASD closure with a reported incidence of 0.01–0.55%; however, it would be higher in the hands of less-experienced operators.[3,5,6] The commonest reasons for ASO dislodgement are the use of an undersized device, large defect size, small LA to accommodate the device, an inadequate or floppy rim of the ASD,[7] persistent device mobility post-implantation, and operator-related technical issues.[8] The most common cause of unsuccessful device deployment is improper patient or device selection.[9] Retrospectively, we found that 14 of our patients had unfavorable septal morphology, which reemphasizes the fact that favorable septal morphology and adequacy of ASD rims are of paramount importance for successful device closure. Transthoracic and TEE play a vital role in the patient as well as device size selection.

The timing of device embolization varies widely from immediate post-deployment period to hospital discharge. Although generally, the incidence of ASO device embolization remains maximum within the first few hours after the procedure, the embolization may also be seen many months after the intervention.[10-12] Device embolization to almost all the chambers of the heart and the great vessels has been reported in the literature.[11-15] In ten of our patients the ASO embolized on to the right side of the heart [Table 3] which includes RA, RV [Figure 1], main pulmonary artery (MPA), junction of MPA with left pulmonary artery (LPA), whereas it embolized to the left side in seven of them, which included the LV [Figure 2], proximal aortic arch, and the descending thoracic aorta (DTA). Following embolization, we noticed that patient remained asymptomatic or presented with variable clinical features, such as changes in heart rate, rhythm, new onset of ectopics, chest pain, and desaturation. These features were dependent on the site of embolization or the degree of obstruction to the blood flow. Fluctuations in the level of systemic oxygen saturation may be a warning feature of device embolization to the PA, which may be associated with varying degrees of obstruction to blood flow.[4] Attempts were made after ASO embolization in our patients to retrieve them percutaneously, although it was associated with a high failure rate, and these attempts resulted in damaging some of the cardiac structures, such as the IVC rim tear we had in a patient. In our study it was found that out of 2237 ASD devices deployed, 17 devices (0.8%) required surgical intervention. In a retrospective analysis, it was found that 1.4% of patients required urgent surgical intervention to retrieve the embolized device.[16]

Figure 1.

Mid esophageal RV inflow outflow view showing ASO device (arrow) abutting the tricuspid valve. ASO = Atrial septal occluder, RA = Right atrium, RV = Right ventricle, LA = Left atrium

Figure 2.

Mid esophageal four-chamber view showing ASD device (arrow) in the left ventricle, abutting the mitral valve. ASD = Atrial septal defect, LA = Left atrium, LV = Left ventricle, RV = Right ventricle

Most of the patients maintain stable hemodynamic parameters after ASO embolization; however, emergency retrieval of the device is needed as it can migrate and lead to complications such as obstruction of blood flow, formation of thrombus on the device, and thromboembolism. In one of our patients, the device migrated from the RA to PA. Device migration beyond the aortic valve may have serious implications as it may compromise the blood supply to the coronary or cerebral circulation. Also, if the device migrates beyond the DTA, it becomes difficult to visualize it on TEE, and retrieval attempts may become unsuccessful. Upon opening the heart in one of our patients, we observed that there was endothelial damage on the atrial septum, and a thrombus had developed on the device surface and on the LA floor, which could risk the development of stroke. We advocate that all patients should be anticoagulated to minimize this risk.

After device embolization, the patients were taken for emergency surgery for device retrieval as well as for closure of the ASD. This was possible through a sternotomy incision in all cases except one, who required a thoracotomy as the device had migrated to the DTA. One patient required a short period of total circulatory arrest (TCA) as the device was high up in the ascending aorta, close to the arch. Although surgical device retrieval is a simple procedure, the surgical team should be prepared to deal with the rapid development of complications associated with the change of position or migration of the device.

Analysis of the US food and drug administration manufacturer database for adverse events documented 17 deaths related to ASO closure of which two were related to device embolization.[12] One of our patients had air embolism, which manifested as sudden desaturation, hypotension, and cardiac arrest. The patient was resuscitated successfully, transferred to the OT and the device was retrieved, which was followed by closure of the ASD. However, when the patient did not wake up in the post-operative period, a brain CT scan was done, which revealed a massive stroke. Incidents of cerebral air embolism post-device closure have been reported;[3,17] however, these patients recovered in due course, but unfortunately our patient did not survive.

The overall risk imposed by device closure of ASD compares favorably with the risk of surgical closure.[18] In spite of ASO device closure being a simple procedure, it carries the risk of device embolization and once this happens the risk may become higher due to subsequent development of complications. In the majority of the cases surgical retrieval is also uneventful, although rarely, there could be certain complications that need the institution of TCA or occurrence of air embolism, which may prove to be fatal as in our case. Once a complication leading to surgery occurs, mortality and morbidity are significantly greater than that of primary surgical ASD closure.[19] The following may help to avoid the complication of device embolization:

• Expertized and skilled interventional cardiologist.

• Proper pre-procedure Echo assessment of size, rims, and location of ASD.

• Before leaving the device in position, it should be confirmed in various positions in the catheter laboratory like Antero posterior (AP), lateral, right anterior oblique (RAO)-cranial, and left anterior oblique (LAO)-cranial. LAO-cranial is like four-chamber view in 2D-ECHO or TEE where we can see the placement of the device on the septum.

• A balloon can be used if the device is in cross position. The size of the balloon depends upon the size of the septum.

• Loop snare should be kept stand by if needed for retrieval of the device.

LIMITATIONS

This was a retrospective analysis from a single center. Larger series are needed from other centers to strengthen our observations and inferences.

CONCLUSION

Device closure of an ASD is an attractive and safe option; however, it may be associated with certain adverse events like device embolization. Once embolization occurs, the device needs to be retrieved either percutaneously or surgically immediately. Most of the surgical extractions are uneventful; however, there could be certain complications that may need repair of valvular apparatus, the institution of TCA, or the need for the lateral position for devices in the descending aorta. Mortality though very rare can occur like the one we had due to complications after air embolization.

Financial support and sponsorship

Nil.

Conflicts of interest

There are no conflicts of interest.