Contraction of the Amplatzer Vascular Plug I and II in Pulmonary Artery and Systemic Venous Collateral Vessels is Safe and of No Hemodynamic or Vascular Consequence in Short- and Mid-term Follow-up

Abstract: Background. The Amplatzer Vascular Plug (AVP) I and AVP II have been used successfully to occlude moderate-large venous collateral vessels (VC) and pulmonary arteriovenous malformations (PAVM) in children and adults with congenital heart disease. Purpose. To report our experience in 4 patients who underwent device occlusion of systemic venous collaterals (3 patients) and device closure of pulmonary arteriovenous malformation (1 patient) that developed contraction of the AVP. Methods. The records of all patients who underwent device occlusion utilizing both AVP I and AVP II between November 2006 and January 2011 were retrospectively reviewed. All available follow-up chest x-rays were reviewed and compared with angiograms obtained post device occlusion. A device 30%-50% larger than the targeted vessel was utilized to occlude the vessel. Results. Four patients were identified with a mean age of 21 years (range, 7 years and 2 months-52 years) and mean weight of 60.6 kg (range, 15.3-131.5 kg). Two patients received AVP I and 2 patients received AVP II. One patient who received the 12 mm AVP I showed moderate contraction. The 3 other patients who received AVP I (6 mm) and AVP II (10 mm, 12 mm) all demonstrated device contraction to the original shape of the device. Mean follow-up time of 24 months (range, 12-40 months) has shown no evidence of hemodynamic or vascular compromise. Conclusions. Short- to mid-term follow-up indicate that contraction of AVP I and AVP II is safe with no evidence of hemodynamic or vascular compromise. Continued long-term follow-up is warranted. 

J INVASIVE CARDIOL 2012;24(4):145-150

Key words: contraction, vascular plug, pediatric interventions, venous collateral, pulmonary arteriovenous malformation

______________________________________________

The Amplatzer Vascular Plug (AVP) I and AVP II (AGA Medical Corporation) have been used successfully to occlude moderate-large venous collateral vessels (VC) and pulmonary arteriovenous malformations (PAVM) in children and adults with congenital heart disease.^1,2 The AVP I is a self-expandable, cylindrical occlusion device made from Nitinol wire mesh that has been in clinical use since 2004,^3,4 The AVP II was approved by the United States Food and Drug Administration for clinical use in late 2006, and is made of a finer, more densely woven Nitinol wire.  AVP I is available in diameters ranging from 4-16 mm and AVP II is available in diameters ranging from 3-22 mm. Both devices have a good track record for safety and versatility and few complications have been noted.^1,2 Recently, we have noticed contraction of the AVP on serial chest x-ray evaluation of patients who have undergone device closure utilizing the AVP. The records of all patients who underwent device occlusion utilizing both AVP I and AVP II between November 2006 and January 2011 were retrospectively reviewed. A total of 4 patients were identified and all available follow-up chest x-rays were reviewed and compared with angiograms obtained post device occlusion. As per the device manufacturer’s recommendation, a device 30-50% larger than the targeted vessel was utilized to occlude the targeted vessel in all cases. The purpose of this case series is to report our experience in 4 patients who underwent device occlusion of systemic VCs (3 patients) and device closure of PAVM (1 patient) who developed contraction of the AVP.

Case 1

A 9-year-old patient with Osler-Weber Rendu syndrome and history of stroke secondary to multiple cerebral arteriovenous malformations (CAVM) presented with systemic desaturation. His past history was significant for embolization of a CAVM at the age of 1 year followed by Gamma knife excision of a second CAVM at the age of 2 years. He remained relatively asymptomatic until he was noted to have a pulse oximetry saturation of 90% while undergoing follow-up magnetic resonance imaging examination of the CAVMs. The patient was referred to the pulmonary service and computed tomography scan of the chest was performed, demonstrating the presence of several PAVMs in both the right and left lung. The patient was then referred to the Pediatric Cardiology service for catheterization and possible device occlusion of the PAVMs.

Cardiac catheterization confirmed the presence of multiple PAVMs. Qp:Qs was 0.8 and the calculated right to left shunt was 0.6 L/min/m2. There was no left to right shunt. Room air saturation was 91%. There was one large left upper-lobe PAVM that was supplied by a feeder vessel measuring between 8.5-9 mm. This was successfully occluded utilizing a single 12 mm AVP II (Figures 1A and 1B). There were two other left lung PAVMs measuring 3-4 mm that were closed with coils and one small right lung PAVM that was left alone. Room air saturation post catheterization increased to 97%.

The patient spent one night in the pediatric intensive care unit and was discharged home the following day. Discharge chest x-ray showed no change in device morphology compared to post-occlusion angiography (Figure 2). The patient was seen for routine follow-up visit at 1 month, 6 months, and 1 year following device closure of PAVM. Device contraction was first noted during review of his chest x-ray obtained at the time of his 1-month visit (Figure 3). Progressive contraction to near its original configuration was noted on chest x-ray films obtained at the time of his 6-month (Figure 4) and 1-year visits (Figure 5). The patient is now 18 months post device closure of the PAVMs and room air saturation measurement remains at 97%. The AVP II remains in stable position without any further evidence of device contraction.

Case 2

A 16-year-old male with tricuspid atresia, transposition of the great vessels S/P Fontan operation presented for routine pediatric cardiology clinic follow-up visit with a history of fatigue. At the time of the clinic visit, he was noted to have a pulse oximetry saturation of 89%. Transthoracic echocardiography demonstrated the presence of a venous collateral arising from the left innominate vein. Because of the above findings, the decision was made to proceed with cardiac catheterization.

Cardiac catheterization demonstrated the presence of a large VC measuring 6.2 mm arising from the left innominate vein and draining into the coronary sinus (source of right to left shunt). The mean pressure recorded in the Fontan circuit was 12 mm Hg. Qp:Qs was 0.8 and the right to left shunt was 0.7 L/min/m². Test occlusion of the VC was performed without increase in pressure within the Fontan circuit, but with increase in saturation measurement to 98%. The VC was successfully occluded with a 10 mm AVP II. The patient spent 1 night in the pediatric intensive care unit and was discharged home the following day with a saturation measurement of 98%.

Discharge chest x-ray showed no change in device morphology compared to post-occlusion angiography (Figure 6). The patient was seen for routine follow-up visit 1 week, 1 month, 3 months, 5 months, 1 year, and 2 years following device closure of the large VC. Device contraction was first noted during review of his chest x-ray obtained at the time of his 1-month visit (Figure 7). Progressive contraction to the original configuration of the device was noted on chest x-ray films obtained at the time of his 5 month visit (Figure 8). The patient is now 2 years post device closure of the large VC and room air saturation measurement remains 98%. The AVP II remains in stable position without any further evidence of device contraction.

Case 3

A 7-year-old patient with tricuspid and pulmonary atresia S/P bidirectional Glenn operation underwent cardiac catheterization prior to completion Fontan operation. At the time of the catheterization, he was found to have a moderate-sized VC measuring 4 mm arising from the left innominate vein and draining into the coronary sinus. In addition. he had long-segment left pulmonary artery stenosis. The mean pressure recorded in the cavopulmonary circuit was 11 mm Hg. The VC was successfully occluded with a 6 mm AVP I and the LPA stent implantation was successfully performed. The patient spent 1 night in the pediatric intensive care unit and was discharged home the following day without any problems. 

Discharge chest x-ray showed no change in device morphology compared to post-occlusion angiography (Figure 9). The patient was then seen for routine follow-up visit at 1 week, 1 month, 3 months, 5 months, 1 year, and 2 years following device closure of the moderate VC. Device contraction to the original configuration was noted at the time of his 3-month visit (Figure 10). The patient is now 3 years and 4 months post device closure of the moderate VC. The AVP I remains in stable position without any further evidence of device contraction.

Case 4

A 52-year-old patient who was initially scheduled to undergo bariatric surgery presented with numbness and tingling of the right side of her body following injection of normal saline to clear a peripheral intravenous line in her left antecubital vein. She was diagnosed with a transient ischemic attack by the neurology service, who then requested that she undergo a transesophageal echocardiogram (TEE) with a bubble study to rule out paradoxical right to left flow through a patent foramen ovale (PFO). The TEE demonstrated microbubbles in the left atrium; however, there was no PFO visualized. She then underwent venography, which demonstrated the presence of a large VC measuring between 7.5 and 10.5 mm arising from the distal left innominate vein draining into the left upper pulmonary vein (source of paradoxical embolism). The large VC was successfully occluded with a 12 mm AVP I. The patient spent 1 night in the adult intensive care unit and was discharged home the following day without any problems. 

Discharge chest x-ray showed no change in device morphology compared to post-occlusion angiography (Figure 11). The patient was seen for routine follow-up visit 1 month, 6 months, and 1 year following device closure of the large left innominate vein to left upper pulmonary vein collateral vessel. At the time of her 1-year follow-up, there was moderate contraction of the device noted (Figure 12). The device remains in stable position without any further change in device morphology.

Discussion

The AVP I and II have now been in clinical use dating back to 2004 for the AVP I and 2006 for the AVP II.  Both devices have a proven track record of safety and few complications have been noted. Because of concerns regarding reconfiguration of both devices to their original size and shape, we reviewed our database and identified 4 patients who underwent device occlusion utilizing either AVP I or II between November 2006 and January 2011. Serial chest x-rays post device occlusion were reviewed. Device contraction was first noted in 2 patients who underwent device closure utilizing AVP II for PAVM and VC at the time of their 1-month visit (Figures 3 and 7), and contraction to the device’s original size and shape were evident at the time of their 6- and 5-month visit, respectively (Figures 4 and 8). The other 2 patients who underwent device closure utilizing AVP I for moderate VC and left innominate vein to pulmonary vein collateral were noted to have contraction to near its original configuration at 3 months and 1 year, respectively. Follow-up exams at a mean time of 24 months (range, 12-40 months) have noted no hemodynamic or vascular compromise as of this writing.      

Contraction of AVP II has previously been reported 5 months after closure of a veno-venous collateral in a patient following bidirectional Glenn operation.⁵ Review of our own series of patients demonstrates that some contraction or reconfiguration of the device starts to occur as early as 1 month post device occlusion. This is true for the 2 patients who underwent occlusion utilizing AVP II. The 2 other patients who received AVP I were noted to have contraction to near its original shape at (1 patient at 3 months and 1 patient at 1 year). We postulate that device contraction in AVP II occurs sooner than AVP I, but that both devices will eventually assume their original configuration by 1 year post device implantation. The reason for the AVP II assuming its original configuration within a shorter time period is most likely secondary to the fact that the Nitinol wires are more tightly and densely woven compared to AVP I. Furthermore, because both devices are made out of Nitinol, which is a shape memory alloy, the natural tendency is for the device to assume its original configuration. Because none of our patients have yet to undergo repeat catheterization, it is difficult to determine if venous regression created by the AVP played a role in contraction of the device. It would be worthwhile to document the angiographic appearance of the occluded vessel should any of the patients require catheterization in the future.  It will also be important to determine if contraction occurs in patients who undergo closure of arterial blood vessels to see if contraction also occurs in arterial blood vessels. With respect to whether or not oversizing of the device played a role in device contraction, the average vessel diameter to device size ratio in the four patients described was 1.4 and none of the devices that were utilized exceeded the device manufacturer’s recommendations. It is important to note that this was a small series of 4 patients and that the longest patient follow-up was 40 months.

In conclusion, short- to mid-term follow-up indicates that contraction of AVP I and II is safe, with no evidence of hemodynamic or vascular compromise. We recommend serial long-term follow-up for all patients who undergo AVP closure and continued reporting of any adverse events should they occur.     

Acknowledgments. The authors would like to thank Sandy Doyle, Vitor Guerra, MD, and Song Yang, MD, for the assistance they provided in preparing and editing the manuscript.

References

1. Hill SL, Hijazi ZM, Hellenbrand WE, Cheatham JP.  Evaluation of the Amplatzer vascular plug for embolization of peripheral vascular malformations associated with congenital heart disease. Catheter Cardiovasc Interv. 2006;67(1):113-119.
2. Schwartz M, Glatz AC, Rome JJ, Gillespie MJ. The Amplatzer vascular plug and Amplatzer vascular plug II for vascular occlusion procedures in 50 patients with congenital cardiovascular disease. Catheter Cardiovasc Interv. 2010:76(3):411-417.  
3. Hill SL, Chisolm JL, Cheatham JP. Initial results of the Amplatzer vascular plug in the treatment of congenital heart disease. Amplatzer Focal Points Newsletter, August 2004. AGA Medical Corporation, Golden Valley, MN.
4. Hijazi ZM. New device for percutaneous closure of aortopulmonary collaterals. Catheter Cardiovasc Interv. 2004;63(4):482-485.
5. Sheridan B, Ward C, Justo R. Reconfiguration of the Amplatzer vascular plug II 5 months after occlusion of venovenous collateral in a bidirectional cavopulmonary circulation. Catheter Cardiovasc Interv. 2010:75(6):857-860. 

______________________________________________

From the ¹Section of Pediatric Cardiology and ²Section of Pediatric Cardiothoracic Surgery, Tulane University, New Orleans, Louisiana.
Disclosure: The authors have completed and returned the ICMJE Form for Disclosure of Potential Conflicts of Interest. The authors report no conflicts of interest regarding the content herein.
Manuscript submitted October 27, 2011, provisional acceptance given November 9, 2011, final version accepted November 22, 2011.
Address for correspondence: Michael R. Recto, MD, Tulane University Section of Pediatric Cardiology, 1430 Tulane Avenue, SL-22, New Orleans LA 70112. Email: mrecto@tulane.edu