Initial Experience with the Amplatzer Vascular Plug IV in Congenital Heart Disease: Coronary Artery Fistula and Aortopulmonary Collateral Artery Embolization

ABSTRACT: Background. A number of percutaneous devices are available to embolize anomalous vessels in congenital heart disease. We report our initial single-center experience with the new Amplatzer Vascular Plug IV (AVP IV) in congenital heart disease to embolize a coronary artery fistula and aortopulmonary collateral arteries in 4 cases. Methods. From August 2009 until April 2010, 7 AVP IV devices were deployed in 4 patients, age range 5 months to 9 years, weight 3.5–27.7 kg. One patient had a large coronary artery fistula, the others had anomalous aortopulmonary collaterals; 2 patients had tetralogy of Fallot with pulmonary atresia, with the other having dextrocardia, anomalous pulmonary venous drainage and pulmonary atresia. Results. In all 4 patients, vessels intended to be closed with the AVP IV were closed successfully with minimal residual shunting and no device failures. Deployed devices ranged from 4–8 mm in diameter. One patient had 4 devices, closing large branching infradiaphragmatic aortopulmonary collaterals. The other 3 patients had single devices. Complete vessel embolization was seen with no device embolization or implantation complication. Conclusion. This preliminary experience illustrates the utility of the AVP IV in congenital heart disease, occluding a coronary artery fistula and aorto-pulmonary collaterals, with devices between 4 mm and 8 mm in diameter, demonstrating its safety and effectiveness. It is particularly useful in embolizing difficult-to-reach large, tortuous vessels with a small-sized catheter in a single procedure. Longer-term follow up in a larger cohort of patients will be required to establish long-term efficacy and device safety. J INVASIVE CARDIOL 2011;23:120–124 Key words: anomalies, coronary and peripheral; congenital heart defects; new devices; percutaneous shunt closure; pediatric interventions ———————————————————— Anomalous arteriovenous connections can be seen in congenital heart disease patients where they may cause significant shunts and abnormal organ perfusion with damaging high-velocity blood flows. Percutaneous embolization is generally the treatment of choice. A number of devices are available for this including coils, occlusion balloons, glue, microspheres and now different novel specific embolization devices or plugs to suit different vascular anatomies. Normally, after a diagnostic catheter has been used, the sheath needs to be changed and a larger guiding catheter is required for their deployment. In addition, with coils and injections, repeated multiple implantations of coils or injections can be required for large-vessel occlusion and there can be a risk of device embolization with peripheral vascular complications. The new Amplatzer Vascular Plug 4 (AVP IV) (AGA Medical Corp., Minneapolis, Minnesota) has been designed to overcome some of these limitations. It is deliverable through smaller diagnostic catheters to reach tortuous distal vessels without the need for catheter change and allows optimal vessel occlusion with a single device. Use of the AVP IV has been described in peripheral vascular embolization such as in bleeding complications following interventional radiology or surgery, but its utility in congenital heart disease in native abnormal vasculature has not been assessed.¹ This is a growing patient population with a need for a minimally invasive effective strategy and a low vascular complication rate. We present our initial single-center experience with this new device in four cases of congenital heart disease, using it to close a coronary artery fistula and aortopulmonary collaterals. The technical aspects of the cases are described, together with the ease of device deployment and initial outcomes.

Methods

From August 2009 to April 2010 four patients were referred for embolization of anomalous vessels: aortopulmonary collaterals or coronary artery fistulae (Table 1). Their age ranged from 5 months to 9 years, and weight 3.5–28.5 kg. Written informed consent was obtained from the parents of the children, with relevant approval from hospital authorities and institutional guidelines. All procedures were performed under fluoroscopy and general anesthesia. Initial assessment included history, clinical examination, chest X-ray, 12-lead electrocardiography (ECG) and two-dimensional (2-D) echocardiography. Inclusion criteria. All patients who were found to have anomalous vessels requiring embolization were considered for device closure with the AVP IV. The decision to deploy the AVP IV was dependent on size and morphology of the vessel and ease of distal access according to the operator. Device. The AVP IV is made of nitinol, a shape memory alloy of nickel and titanium. It has a multilayered, flexible fine meshwork construction which self-expands when deployed, its double-lobed design assisting in blocking the target vessel and its floppy distal section enabling the device to be steered through tortuous vessel anatomy. The plug is softer than the other AVPs due to thinner wires used in its construction and this, together with its bi-lobed design, allows it to be delivered via catheter in tortuous vessels. This design also prevents the catheter from straightening during deployment, assisting catheter positioning. It has a radiopaque marker band at each end with a microscrew attachment at one end that attaches it to a 155-cm delivery wire. Once deployed, the device is magnetic resonance imaging (MRI)-compatible at a strength of 1.5 T. The device is packaged preloaded in a specific loader and is deliverable in a 0.038 inch guidewire-compatible diagnostic catheter (e.g., Cordis^® TEMPO^® 4 Fr or Boston Scientific IMAGER II^™ 5 Fr). The AVP IV is available in a range of diameters from 4–8 mm in 1-mm increments with lengths of 10–13.5 mm, and can be deployed, recaptured and redeployed to assist secure placement. Procedure. Femoral arterial access was obtained with 4 French (Fr) sheaths in the youngest cases and a 6 Fr sheath in the oldest. Heparin was administered at a dose of 100 IU/kg. Aortograms were performed initially with a pigtail catheter in the aorta to visualize the anomalous vessel position and approximate its size before selective catheterization with a 0.038 inch guidewire-compatible 4 Fr diagnostic catheter to enable more accurate assessment for sizing. A device 30–50% larger in diameter than the target vessel diameter was generally selected for closure as suggested by the manufacturer. In each of our cases this was done by comparing the size of the target vessel to the catheter size as standard. The selected AVP IV device was inserted either via the same diagnostic catheter or a Mullins sheath. The loader has a tapered tip to guide insertion into the delivery catheter, insertion through a hemostasis valve or Y-connector not being recommended. The loading mechanism has a stopcock at the distal end, opened briefly to allow blood backflow and ensure air is purged from the delivery system. At the site of intended deployment the delivery catheter was slowly retracted to leave the device within the vessel. Device position was easily verified by its radiopaque marker bands (Figures 1–3). If device position was unsuitable it could be recaptured by stabilizing the delivery wire and readvancing the delivery catheter. The device was released from the delivery wire in each case by rotating the wire in a clockwise direction until the device separated under fluoroscopic screening. Repeat aortography was performed several minutes following release to check for effective embolization. Periprocedural antibiotics were given according to institution protocol. All patients were discharged 24–48 hours later, following clinical examination, 12-lead ECG, chest X-ray and a satisfactory echocardiogram.

Results

Patient characteristics, specific anomalous vessels and devices used are shown in Table 1. The devices used included 4, 5, 6 and 8 mm diameter sizes. A coronary artery fistula arising from the proximal right coronary artery to right ventricle and aortopulmonary collaterals arising from the descending aorta and the subclavian arteries were treated. All intended vessels were occluded with the AVP IV with no deployment failures. The first patient had a large coronary artery fistula arising from the proximal part of the right coronary artery leading into the right ventricle (Figure 1); it was recognized by echocardiography following cardiology review for a murmur. There was evidence of a significant shunt with right heart dilatation. Initial angiography was performed with a 4 Fr multipurpose catheter, revealing a large ectatic fistula (Figure 1A). A 0.018 inch Terumo guidewire (Terumo Medical Corp., Somerset, New Jersey) was manipulated through the vessel, into the right ventricle and then into the main pulmonary artery. Here it was captured with a snare catheter and exteriorized via the right femoral artery to form an arteriovenous loop. It was thought that maintaining this arteriovenous loop would be safest during device deployment, allowing continued access to the anomalous vessel until final device release and attempted closure. A 5 Fr Mullins sheath was inserted via the femoral venous access and moved along the wire on its dilator to the right atrium, right ventricle and then the coronary fistula. A 6 mm AVP IV device was then delivered via the Mullins sheath, keeping the wire in position (Figures 1B and 1C). As the position appeared stable, the guidewire was removed and the device released. Within 2 minutes, successful vessel embolization was observed (Figure 1D). The AVP IV seemed ideally suited for fistula occlusion due to the size of the target vessel and the tortuosity required to get to it. Patient no. 2 had systemic pulmonary artery collaterals arising from the infradiaphragmatic descending aorta towards the right inferior pulmonary lobe (Figure 2). Pressure in the aorta was 75/44, mean 58 mmHg, with pressure in the largest collateral of 42/29, mean 35 mmHg. The largest major aortopulmonary collateral artery (MAPCA) was closed with an 8 mm AVP IV with the smaller one closed with a 5 mm AVP IV initially (Figure 2B). Follow-up angiography 12 minutes later still suggested some residual flow in both vessels, so additional 5 mm and 4 mm devices were deployed in each MAPCA, respectively (Figure 2C). Angiography with a Berman catheter after a further 20 minutes suggested that the MAPCAs were successfully occluded (Figure 2D). The use of coils would have required multiple devices and, using the AVP IV, the vessels were accessed off the aorta with a 5 Fr multipurpose catheter and subsequent embolization without the need for further catheter exchange and guidewire. The third patient had 3 collaterals embolized. The first collateral divided into two branches, one towards the right middle and upper lobes and another towards the left superior lobe. The left-sided vessel was closed with 2 coils (both 3 x 3 m) and the right-sided branch was occluded with the AVP IV, as decided by the operator. The AVP IV in this case allowed rapid embolization of a large collateral without the need for multiple devices and was deliverable using the same multipurpose catheter used for vessel angiography. A collateral vessel from the abdominal aorta to the right middle lobe was embolized by a 5 x 5 mm coil, and the final collateral from the right subclavian to the right upper and middle lobes was closed with two 5 x 3 mm coils. The proximal right pulmonary artery was then balloon-dilated with further implantation of a Jomed 28 mm stent (Abbott Vascular, Abbott Park, Illinois) on a 10 x 30 mm balloon as an additional procedure. A Palmaz-Genesis 29 mm stent (Cordis Corp., Miami Lakes, Florida) was then implanted on a 12 x 30 mm balloon and then a further AndraStent 30 XXL (Andramed GmbH, Reutlingen, Germany) on a 15 x 40 mm balloon to obtain optimal dilatation of the left pulmonary artery. Angiography post deployment of the AVP IV confirmed aortopulmonary collateral embolization. The fourth patient, with previous tetralogy of Fallot repair, was undergoing investigation for high pulmonary artery pressures. In addition to left pulmonary artery stenosis, a tortuous MAPCA was found arising from the right subclavian artery (Figure 3). After dilatation of the pulmonary artery and stent implantation, the MAPCA was intubated with a 4 Fr catheter for accurate visualization. Due to the difficulty in access and tortuous course, the vessel was thought suitable for the AVP IV, obviating the need for further catheter exchange. Its floppy distal tip and thin nitinol wire construction assisted in guiding the device to the target site around the tortuosity of the transverse aorta, right innominate and subclavian artery without straightening the catheter. A single 6 mm AVP IV was delivered with successful embolization (Figures 3C and 3D). The use of another vascular plug would have required a catheter change and further manipulation to regain access to the target vessel, increasing procedure time and possibly catheter-related morbidity with larger access, contrast injection and fluoroscopy required.

Discussion

Occlusion of abnormal vessels is often required in congenital heart disease. A device that permits a variety of vessel widths to be occluded effectively, can be deliverable by small-French catheters without the need for catheter exchange through tortuous vessels, and does not require consideration of multiple or additional devices — is helpful in reducing procedural times and vascular comorbidity. Abnormal vessels that may require embolization include coronary artery fistulas and MAPCAs. Coronary artery fistulas connect a coronary artery to one of the cardiac chambers or great vessel, and early treatment is generally recommended to prevent potential heart failure, myocardial ischemia, risk of infective endocarditis, pulmonary hypertension and possible aneurysmal vessel rupture.^2,3 In turn, MAPCAS can be defined as the presence of any arterial vessels, apart from the ductus arteriosus, that may compensate for obstructed flow from the right ventricle to the pulmonary arteries.⁴ They may cause problems on cardiopulmonary bypass with flow from collaterals into the left atrium causing blood in the surgical field, and high pulmonary blood flow and cardiac volume overload can cause difficulty in weaning off ventilation postoperatively. The first reports of the use of embolization techniques in congenital heart disease was for aortopulmonary collateral artery embolization in the early 1980s.⁵ Initial techniques utilized latex and silicone balloons, gel foam fragments and steel coils. Erroneous embolization was a risk, and these devices could be difficult to place in distal tortuous vessels; multiple devices were often required to provide adequate vessel closure. The Amplatzer^® vascular plug (AVP) family of devices were devised to overcome some of these limitations.⁶ Use of the AVP I was first described in 2004;⁷ it was designed to allow embolization of peripheral veins and arteries with a single device like a “cork in a bottle” to allow controlled deployment with a variety of cross-sectional vessel coverage. Like the entire AVP family, it consists of a distensible nitinol mesh that has “shape memory” and is available in a range of sizes from 4–18 mm in 2-mm increments, the smaller devices being 7 mm in length and the largest 8 mm. Generally, a device 30–50% larger in diameter than the target vessel to be embolized is selected. However, this first device is deployed with a guide catheter that ranges from 5 Fr to 8 Fr in size. The AVP II obtained FDA approval and CE mark in 2007 and has a multisegmented, multilayered design with a faster vessel occlusion time and 3 lobes that give it different vessel conformational properties than the AVP I. Similarly, it is available in a variety of diameters from 3–28 mm in 2-mm increments, with device lengths varying from 6–9 mm depending on size. Similarly, a device size 30–50% larger in diameter at the site of intended vessel occlusion is recommended and a guiding catheter of at least 5 Fr is required, as the device is deployed over a guidewire. The AVP III obtained European CE mark in 2008 and features an oval shape and enlarged rims compared to the AVP II, which may improve wall opposition in high-flow vessels. It is also recommended that a device 30–70% larger than the target vessel be used, and it comes in a 4 x 2 mm waist size up to 14 x 5 mm, with the rims being 2 mm larger than the waist. Again a guide catheter is required, varying in size from 6–9 Fr. The AVP IV is the latest of the vascular plug family, obtaining European CE mark in July 2009, and is uniquely deliverable by a 4 Fr catheter. Exchange for a guide catheter is not required, thus minimizing the number of steps required for embolization and allowing more tortuous distal vessels to be reached with its smaller delivery system. The AVP IV has a floppy distal section at the end of its delivery wire that assists in steering it through tortuous vessels and, promoting rapid embolization, it has a multilayered double-lobed design. Thin nitinol wire is used in its construction, which also helps it to be floppy and not straighten the delivery catheter during deployment. It is available in 4–8 mm diameter sizes in 1-mm increments, with lengths of 10–13.5 mm and, unlike the other vascular plugs, all of the AVP IVs can be deployed without a guidewire. Little has been published comparing these devices. One study examined the utility of the AVP I and AVP II in patients with congenital heart disease, finding that the AVP II may be better in high-flow tubular structures and patients with type C patent ductus arteriosus.⁸ All target lesions were successfully occluded irrespective of device used, however. Definitive trials comparing the utility of the different AVP devices and to other embolization methods will be difficult due to the heterogeneity of lesions treated and the need to find large numbers of patients with similar pathologies. The AVP IV appeared to have the advantage over the other vascular plugs in that it can close large anomalous vessels with deployment of a single device and can reach tortuous distal vessels due to its deliverability via a smaller sheath. As a guidewire and new delivery catheter are not needed, it requires less manipulation, reducing fluoroscopy and procedural times, and the smaller catheters may minimize vascular access morbidity. Compared to coils though, a range of smaller devices may be helpful and it will be more expensive, at least initially. If a target vessel is particularly difficult to enter, then the AVP IV will have the advantage in that a single device, rather than multiple coils, will be required. There may be concerns about the safety of nitinol devices with regard to erosion at particular sites over the longer term. Nothing has been reported regarding the other AVPs, but long-term follow up is generally advocated. We present our initial single-center experience with this new device in patients with congenital heart disease, describing the first cases in which it was used to close a coronary artery fistula and aortopulmonary collateral vessels. The AVP IV was deliverable through small-French catheters to tortuous distal vessels and, for 3 of the cases, did not require catheter exchange, which would have been necessary with the other vascular plugs. In one case, an additional device was required in each target vessel despite the largest size being used, with rapid occlusion occurring after the second device deployment. This new device was thus successfully used without complication, with minimal manipulation in a number of different target vessels and appears to be a helpful adjunct to the embolization armamentarium. It will be useful in large tortuous or difficult-to-enter target vessels compared to other AVP models and other embolization devices. Initial fluoroscopy and follow up suggest that the AVP IV was particularly effective, but studies will be required to identify any long-term effects.

References

1. Ferro C, Rossi UG, Bovio G, et al. The AMPLATZER Vascular Plug 4: Preliminary Experience. Cardiovasc Intervent Radiol 2009;33:844–848.
2. Liberthson RR, Sagar K, Berkoben JP, et al. Congenital coronary arteriovenous fistula. Report of 13 patients, review of the literature and delineation of management. Circulation 1979;59:849–854.
3. Said SA, Lam J, van der Werf T. Solitary coronary artery fistulas: A congenital anomaly in children and adults. A contemporary review. Congenit Heart Dis 2006;1:63–76.
4. McGoon MD, Fulton RE, Davis GD, et al. Systemic collateral and pulmonary artery stenosis in patients with congenital pulmonary valve atresia and ventricular septal defect. Circulation 1977;56:473–479.
5. Reidy JF, Jones OD, Tynan MJ, et al. Embolization procedures in congenital heart disease. Br Heart J 1985;54:184–192.
6. Amplatzer vascular plug family AGA Medical Corporation, 2010. URL:https://www. international.amplatzer.com/InternationalProducts/VascularPlugFamily/tabid/528/Default.aspx [July 2010].
7. Hijazi ZM. New device for percutaneous closure of aortopulmonary collaterals. Catheter Cardiovasc Interv 2004;63:482–485.
8. 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;76:411–417.

———————————————————— From the Department of Pediatric Cardiology and Adult Congenital Heart Disease, Policlinico San Donato IRCCS, Via Morandi 30, 20097 San Donato Milanese (MI), Italy. The authors report no conflicts of interest regarding the content herein. Manuscript submitted June 8, 2010, provisional acceptance given July 6, 2010, final version accepted September 27, 2010. Address for correspondence: Dr. Gianfranco Butera, Department of Paediatric Cardiology and Adult Congenital Heart Disease, Policlinico San Donato IRCCS, Via Morandi 30, 20097 San Donato Milanese (MI), Italy. E-mail: gianfra.but@lycos.com