Since the initial description by Portsmann et al of nonsurgical closure of patent ductus arteriosus (PDA),1 the percutaneous approach has developed into standard clinical practice. Coils,2,3 and subsequently devices, 4,5 have broadened the range of patients and anatomies treatable with percutaneous techniques. In spite of the range of devices and coils available for this procedure, there remains a limitation in small children with moderate-sized shunts and larger children with small PDAs. The recently released Amplatzer Ductal Occluder 2 (ADO2) (AGA Medical Corp., Plymouth, Minnesota) was designed specifically for these indications. We report our initial experience with this device.
MethodsPopulation. Patients presenting with a clinically significant PDA to a single general hospital in Jerusalem, Israel, were referred for percutaneous closure with the ADO2. In instances where the PDA was larger than 5.5 mm in diameter, use of the device was not considered. Patients with additional cardiac anomalies requiring surgical correction were referred for surgery, as were patients Device and delivery. The ADO2 device is a modification of the ADO1 device as produced by AGA Medical Corporation. It is one of the range of vascular occlusion devices based on nitinol wire meshes shaped in sequential lobes. The ADO2 is characterized by two low-profile retention discs for placement in the aorta and pulmonary artery (PA) (Figure 1). A connecting waist of variable diameter and length for positioning within the PDA itself is attached to the retaining discs by two articulations. This allows the positioned, relatively soft device to adapt to the patient’s anatomy rather than distorting the anatomy to its shape as do the ADO1 and other plugs.6 When correctly positioned, there are six layers of nitinol mesh interrupting flow, allowing for rapid occlusion in the absence of integrated fabric patches. The sizing of the device is based upon the waist diameter and length measured by angiography. Diameter ranges are 3–6 mm and lengths are available in 4 or 6 mm. The retaining discs have a diameter 6 mm greater than the waist and the device is symmetrical with both retaining discs being of equal size (Figure 1). This allows for either transvenous or transaortic deployment. Devices with waists of 3 and 4 mm can be delivered through a 4 French (Fr) delivery system, while the larger devices require a 5 Fr system. The delivery system comprises a braided, flexible and tapered delivery sheath with a 90° curve and a flexible delivery cable. Procedure. Devices were deployed using either transpulmonary or transaortic approaches. The standard transpulmonary technique for device deployment has been described for both the ADO14 and ADO2.7,8 Briefly, the duct is crossed from the pulmonary artery, the aortic retainer is then exposed in the main aorta and the delivery system withdrawn to allow for the waist to be deployed in the duct itself. The pulmonary retainer is then exposed to abut against the pulmonary arterial wall. Device positioning is confirmed by aortography using a pigtail catheter positioned in the aorta. An alternative and more rapid approach is direct passage through the duct from the aorta, subsequent exposure of the pulmonary disc, followed by deployment of the device waist and then aortic disc exposure for final positioning. Aortographic contrast injections are performed through the delivery sheath to confirm device positioning. For PDAs of Results Seven patients with appropriate PDAs were treated. The patient age range was 7 months to 23 years (median 3 years 4 months), with a weight range of 7–56 kg (median 12 kg). A single adult (23 years) was treated under local anesthetic, with all remaining cases performed under general anesthetic. Patient demographic and intraprocedural characteristics as well as follow-up results are presented in Table 1. Device deployment was successful in all cases, and intraprocedural ductal closure was successful in all cases. Next-day echocardiography demonstrated ductal occlusion in 6/7 cases with a small, silent, but significant, shunt demonstrated by echocardiography in a single case. At 30 days, echocardiography confirmed ductal occlusion without need for further intervention in all cases. The device was deployed using a transvenous approach in 6 cases (Figure 2). In 1 case, a transaortic approach was used (Figure 3). In an additional case, a transaortic approach was attempted and the device was well-positioned on the pulmonary side and in the duct itself. However, due to a relatively small and asymmetric ductal ampulla, the aortic lobe’s inferior aspect fell into the ampulla, resulting in extension and protrusion of the superior aspect of the aortic lobe into the aorta. This could not be corrected despite multiple maneuvers including recapture and redeployment from different angles. The device was removed and redeployed from a transvenous approach, allowing for improved aortic positioning and successful device deployment (Figure 4). In 2 cases where the duct was very short and the ampulla large, the device deployed from the transpulmonary approach resulted in the aortic lobe falling into the ampulla, with its articulation crossing the duct and the middle lobe and the pulmonary disc protruding into the pulmonary artery. Due to the short length of the ducts, the devices were able to be placed with only one or other of the articulations in the duct. The pulmonary protrusion was considered preferable and the devices were released with excellent acute results and no gradient around the device in the pulmonary artery. One of these 2 cases resulted in a recurrent shunt detected the next day. This complex case warrants more extensive description. A 7-month (7 kg) female presented with a moderate-sized PDA that was sized 3 mm at the waist by echocardiography and 1.8 mm by angiography. Due to the echocardiographic sizing, a 4 x 4 mm device was deployed using a transvenous approach. The duct was very short and the waist positioned on the pulmonary side with a residual leak. The device was repositioned with the waist in the duct; complete closure was achieved, however, due to the angle that the aortic disc assumed with superior lobe protrusion into the aorta (Figure 5) and significant turbulence detected by echocardiography. With a small gradient in the aorta at the device site, it was considered prudent to remove the device and replace it with a smaller one. A 3/4 device was then deployed with the waist in this instance also assuming position in the pulmonary artery. The aortic disc sat well in the ampulla with no aortic protrusion; there was no significant turbulence in the pulmonary artery and the duct was occluded with no residual flow detected by echocardiography or angiography. This was considered a satisfactory position and the device was then released with a good result. Next-day echocardiography demonstrated a recurrent or residual shunt predominantly through the device which was still well positioned. The shunt was relatively small and there was clear evidence of hemodynamic benefit of the procedure with reduction of left ventricular size and mitral regurgitation. Echocardiographic follow up at 7 days confirmed the residual shunt with no significant difference and no evidence of hemolysis. At the 30-day echocardiographic follow up, there had been complete resolution of the shunt with an excellent final result.
DiscussionThe ADO2 device provides a solution for the closure of small- to moderate-sized PDAs as well as large PDAs in small children. The ability to position the device such that each of the retention discs assumes its independent orientation significantly reduces the risk of protrusion, anatomical distortion and displacement. The multiple layers of nitinol (two for each of the three lobes) allow for rapid ductal occlusion in the absence of integrated patches. This reduces residual shunt risks and the need to deploy additional coils or devices and results in an extremely low-profile device during deployment. Furthermore, the transaortic approach has significant benefit, allowing the procedure to be extremely simple and rapid with the use of a single arterial puncture (often 4 Fr) and the ability to perform test injections through the delivery sheath. This is a feature that, although not exclusive to this device, certainly adds to its deployment flexibility. Several important issues emanate from this initial experience. The first is the tendency of the aortic retention disc to stretch such that, particularly if the aortic ampulla is large and the duct short, the central waist may protrude into the PA, leaving the PA retaining disc redundant. This occurs even more so if the device is significantly oversized. In our experience, this positioning was not problematic and no gradient was observed. However, in a previous report,7 2 cases with 15 mmHg pulmonary arterial gradients were noted. As such, the shorter device (4 mm) will usually be applicable unless the duct is extremely long. Recently, the benefits of this device in extremely long ducts have been highlighted,9 and this is clearly its strength. Conversely, this may require additional care with short ducts. The length of the duct is routinely measured as a straight line from an extrapolated line across the mouth of the aortic ampulla to an extrapolated line across the pulmonary arterial entrance of the duct (Figure 6).9 The manufacturer recommends that when this length is > 5 mm, the 6 mm-long device should be used.8 This measured length, although anatomically correct, does not truly represent the length of duct to be occluded. This emanates from the discrepancy between the minimal ductal diameter and the often much larger aortic ampulla ostial diameter as we (Table) and others 8 have shown, particularly in PDAs of Krichenko Types A and E (funnel types).10 As the aortic disc is 6 mm greater in diameter than the nominal waist diameter, we propose that the device length should be chosen based on a length measurement taken from the pulmonary arterial entrance of the duct to a point in the ampulla where the diameter measures the device waist plus 6 mm (the selected devices aortic disc diameter) (Figure 7). Using this length, the majority of devices will be shorter (4 mm), rather than longer. In our limited experience, all ducts were well closed with the shorter device, although according to the manufacturer’s recommendation, a longer device should have been chosen. In the single case in our series where a longer device was selected, there was some device redundancy in the pulmonary artery. This device selection method may reduce the risk of device migration, as has been reported with this device,8 possibly due to the less snug fit of the aortic disc in the ampulla. The ADO2 device was designed by the manufacturers with the express intention that the retaining discs would be positioned in the aorta and pulmonary artery. However, in the presence of a large ampulla, the aortic retaining disc will be positioned within the ampulla and not the aorta, as described. This scenario of a large ductal ampulla aortic ostium is also important when deployment is performed from the aortic side. In such situations, the device can be well positioned in the duct, but with the aortic retaining disc falling partially (usually inferiorly) into the ampulla, leaving the superior aspect projecting perpendicularly into the aorta. This may not be correctable using the transaortic approach, as there is an extremely limited ability to push the device into the ampulla without deflecting it superiorly. This can further worsen the device’s position and may be traumatic to the duct itself, resulting in spasm or even rupture. In such instances, however, when approached transvenously, the aortic disc can be well positioned in the ampulla, while still allowing the rest of the device to assume its intended position.
ConclusionThe ADO2 broadens the spectrum of PDAs that can be simply and safely treated with devices. Due to the range of anatomy and sizes that are treated, a single device will always be limited as a total solution. However, the flexibility of the articulations, coupled with the alternative deployment techniques, allows this device to simplify the treatment in a range of small patients and specific ductal anatomies that are more challenging. In extremely short ducts or where the ampulla is very large compared to the aortic disc, the device may result in protrusion on either side. An alternative length measurement of the duct length to accommodate the device may improve device length selection. Broader experience is required to further delineate ADO2 device and patient selection, as well as to document its long-term efficacy and safety.
1. Porstmann W, Wierny L, Warnke H. Closure of persistent ductus arteriosus without thoracotomy. Ger Med Mon 1967;12:259–261.
2. Cambier PA, Kirby WC, Wortham DC, Moore JW. Percutaneous closure of the small (less than 2.5 mm) patent ductus arteriosus using coil embolization. Am J Cardiol 1992;69:815–816.
3. Rosenthal E, Qureshi SA, Reidy J, et al. Evolving use of embolisation coils for occlusion of the arterial duct. Heart 1996;76:525–530.
4. Pass RH, Hijazi Z, Hsu DT, et al. Multicenter USA Amplatzer patent ductus arteriosus occlusion device trial: Initial and one-year results. J Am Coll Cardiol 2004;44:513–519.
5. Gudausky TM, Hirsch R, Khoury PR, Beekman RH 3rd. Comparison of two transcatheter device strategies for occlusion of the patent ductus arteriosus. Catheter Cardiovasc Interv 2008;72:675–680.
6. Gruenstein DH, Bass JL. Experimental evaluation of a new articulated Amplatzer® ductal occluder without fabric. Catheter Cardiovasc Interv 2009;74:482–487.
7. Thanopoulos B, Eleftherakis N, Tzannos K, Stefanadis C. Transcatheter closure of the patent ductus arteriosus using the new Amplatzer duct occluder: Initial clinical applications in children. Am Heart J 2008;156:917.e1–917.e6
8. Forsey J, Kenny D, Morgan G, et al. Early clinical experience with the new amplatzer ductal occluder II for closure of the persistent arterial duct. Catheter Cardiovasc Interv 2009 Mar 9 (Epub ahead of print).
9. Morgan G, Tometzki AJ, Martin RP. Transcatheter closure of long tubular patent arterial ducts: The Amplatzer Duct Occluder II-A new and valuable tool. Catheter Cardiovasc Interv 2009;73:576–580.
10. Krichenko A, Benson LN, Burrows P, et al. Angiographic classification of the isolated, persistently patent ductus arteriosus and implications for percutaneous catheter occlusion. Am J Cardiol 1989;63:877–880.
From the Department of Cardiology, Shaare Zedek Medical Center, Jerusalem, Israel. The authors report no financial relationships or conflicts of interest regarding the content herein. Manuscript submitted October 22, 2009 and accepted November 4, 2009. Address for correspondence: David Meerkin, MBBS, Director of Experimental Cardiology, Shaare Zedek Medical Center, POB 3235, Jerusalem, Israel, 91031. E-mail: meerkin@SZMC.org.il