Author + information
- Received June 4, 1999
- Revision received October 15, 1999
- Accepted November 19, 1999
- Published online March 15, 2000.
- Basil (Vasilios) D Thanopoulos, MD∗,*,
- Fakhri A Hakim, MD†,
- Aktham Hiari, MD†,
- Yousef Goussous, MD†,
- Evangelia Basta, MD∗,
- Armine A Zarayelyan, MD∗ and
- George S Tsaousis, MD∗
- ↵*Reprint requests and correspondence: Dr. Basil (Vasilios) D. Thanopoulos, Department of Pediatric Cardiology, “Aghia Sophia” Children’s Hospital, Thivon & Levadias Street, Athens 11527, Greece
The aim of this study was to report further experience with transcatheter closure of the patent ductus arteriosus (PDA) using the Amplatzer duct occluder (ADO).
The design of previously used devices is not ideal for this purpose, and their use has been associated with several drawbacks, especially in large PDAs.
Forty-three patients, aged 0.3 to 33 years (mean 6.4 ± 6.7 years), with a moderate to large, type A to E PDA, underwent attempted transcatheter closure using the ADO. The device is a plug-shaped repositionable occluder made of 0.004-in. nitinol wire mesh. It is delivered through a 5F to 6F long sheath. The mean PDA diameter (at the pulmonary end) was 3.9 ± 1.2 mm (range 2.2 to 8 mm). All patients had color flow echocardiographic follow-up (6 to 24 months) at 24 h, 1 and 3 months after closure, and at 6-month intervals thereafter.
The mean ADO diameter was 6.1 ± 1.4 mm (range 4 to 10 mm). Complete angiographic closure was seen in 40 of 43 patients (93%; 95% confidence interval [CI] 85.4% to 100%). The remaining three patients had a trivial angiographic shunt through the ADO. At 24 h, color flow mapping revealed no shunt in all patients. A 9F long sheath was required for repositioning of a misplaced 8-mm device into the pulmonary artery. The mean fluoroscopy time was 7.9 ± 1.6 min (range 4.6 to 12 min). There were no complications. No obstruction of the descending aorta or the pulmonary artery branches was noted on Doppler follow-up studies. Neither thromboembolization nor hemolysis or device failure was encountered.
Transcatheter closure using the ADO is an effective and safe therapy for the majority of patients with patency of the arterial duct. Further studies are required to establish long-term results in a larger patient population.
In 1967, Porstmann et al. (1)were the first to employ successfully an Ivalon plug for transcatheter closure of a patent ductus arteriosus (PDA). Since then, various occluding devices and coils have been used in numerous patients with quite satisfactory results. In spite of obvious advantages of the percutaneous techniques in infants and small children with moderate and large PDAs, the percutaneous techniques have significant limitations. Large delivery sheaths, cumbersome implantation techniques, device embolizations and a high rate of residual shunting are some of the drawbacks of previously prescribed techniques (2–13). Recently, Masura et al. (14)reported complete closure of moderate-to-large sized PDAs in 23 of 24 patients using the Amplatzer duct occluder (ADO), a new device manufactured by AGA Medical Corporation and experimentally tested at the University of Minnesota. In this report, we present further experience in 43 patients, with ductus measuring up to 8.0 mm in diameter, who underwent transcatheter closure with the ADO.
From May 1997 to February 1999, 43 patients (14 males and 29 females) with PDA underwent attempted transcatheter closure using the ADO. The median age of the 43 patients was 6.6 years (range 0.3 to 33 years) and the median weight was 20.3 kg (range 4.2 to 77 kg). Ten patients had symptoms of heart failure and/or failure to thrive. All patients had echocardiographic evidence of significant left-to-right shunting through the PDA with left atrial enlargement and ventricular volume overload. Informed parental consent was obtained in each patient. Patients with additional congenital cardiac anomalies that would require cardiac surgery, body weight less than 4 kg, and PDA diameter smaller than 2 mm were excluded from the study.
Device and delivery system
The ADO (AGA Medical Corporation, Golden Valley, Minnesota) has been described in detail in previous reports (14). In brief, this occluder is a cone-shaped device 7 mm in length made of a 0.004-inch Nitinol wire mesh (Fig. 1). A 2-mm retention skirt extends radially around the distal part of the device assuring secure fixation in the mouth of the PDA. Prostheses are currently available in sizes ranging from 6-4 to 14-12 mm at increments of 2 mm. The larger measurement is at the aortic side and the smaller diameter is at the pulmonary end. The device is attached by a recessed microscrew to a 0.038-in. delivery cable made of stainless steel. It is delivered through a 5F to 6F long sheath. For introduction into the delivery sheath, the device is pulled into a special Teflon loader.
The technique of transcatheter closure was similar to that described by Masura et al. (14). After percutaneous puncture of the femoral vein and artery, a complete hemodynamic evaluation was performed with pressure and saturation measurements taken in all cardiac chambers. A biplane descending aortogram in anterioposterior and lateral projection was performed with a 4F or 5F pigtail catheter, which was introduced percutaneously, to define the size and anatomy of the ductus. In four patients (patients 2, 31, 37 and 42), it was difficult to measure the diameter of the PDA by aortography. Therefore, “balloon sizing” was performed from the venous side with a balloon-tipped end-hole catheter (Arrow). Subsequently, a 4F or 5F cobra type I (Cordis) catheter was advanced percutaneously from the venous side through the PDA into the descending aorta. Using an exchange 260-cm 0.035-in. guidewire, the cobra catheter was exchanged for a 5F or 6F delivery sheath that was advanced directly through the femoral vein and positioned in the proximal aorta.
A proper size occluder (smaller diameter 1 to 3 mm larger than the diameter of the pulmonary end of the ductus or equal to the measured balloon “stretched” diameter) was screwed to the delivery cable, compressed into the loader and introduced into the guiding sheath. Under fluoroscopic guidance, the ADO was advanced into the descending aorta, where the retention disc was deployed. Once in position, the disc was pulled gently against the orifice of the PDA, which could be felt as a rhythmic tugging sensation in synchrony with the cardiac cycle. Correct position was confirmed by a hand injection of contrast medium through the aortic catheter into the descending aorta. Using gentle tension on the delivery cable, the sheath was pulled back to deploy the conical part of the device into the ductus. Two-dimensional color Doppler echocardiography and descending aortography were performed after device placement to document residual shunts, and left pulmonary artery or aortic obstruction. Once optimal position was confirmed, the ADO was released by counterclockwise rotation of the delivery cable. A repeat aortogram was performed to check for residual shunts. Prophylactic antibiotics were not administered during the procedure. All patients were discharged 24 h after the procedure on no medications. Endocarditis prophylaxis was discontinued at the six-month follow-up visit if the ductus was completely closed.
A chest radiograph and complete two-dimensional and color Doppler echocardiographic studies were performed on all patients at 24 h, 1 month and serially at 3- to 6-month intervals. Special attention was paid to residual ductal flow, left pulmonary artery or aortic stenosis and wire fractures.
Results are expressed as mean value ± SD, with confidence intervals given where applicable.
The clinical data and the outcome for the 43 patients are shown in Table 1. According to the alphanumeric classification proposed by Krichenko et al. (15), 32 patients had a PDA type A, four had PDA type B, three had PDA type C, two had PDA type D and two had PDA type E. Delivery of the device was successful in all patients. The mean PDA minimal (pulmonary end) diameter determined by aortography was 3.9 ± 1.2 mm (range 2.2 to 8 mm). The mean ADO smallest diameter was 6.1 ± 1.4 mm (range 4 to 10 mm). There was left-to-right shunting across the PDAs in all patients, both by oxymetry and angiography. The pulmonary/systemic flow ratio (Qp/Qs) varied between 1.5 and 4 (mean 2.4 ± 0.7). Two patients (patients 36 and 41) with significant pulmonary artery hypertension (mean pulmonary artery pressure of 43 and 48 mm Hg, respectively) showed immediate improvement with a reduction of the mean pulmonary artery pressure to 22 and 28 mm, respectively.
In one patient (patient 37), during the deployment of the conical part, an 8-mm device slipped into the pulmonary artery. Several efforts to retract the device into the 6F delivery sheath failed. The device being tilted by the blood flow into an almost vertical position in relation to the sheath made recapturing impossible. A 9F Amplatzer delivery sheath had to be introduced, which allowed recapturing and successful redeployment of the device. In one patient (patient 31), a small infant with a large tubular (type C) PDA, deployment of the retention disc against the orifice of the ductus caused severe aortic obstruction. Therefore, both the retention disc and the conical part of the device were deployed within the ductus (Fig. 2).
Complete angiographic closure was present in 40 of 43 patients (93%; 95% confidence interval [CI] 85.4% to 100%) (Fig. 3). Foaming (trace angiographic residual shunt and no contrast jet) was seen in 10 patients (28%) but disappeared within 15 min. In three patients, there was a trivial residual shunt immediately after the procedure. Fluoroscopy time was 7.9 ± 1.6 min (range 4.6 to 18 min), and total procedure time was 21.8 ± 5.8 min (range 15 to 48 min). No complications were observed. All patients had good distal lower extremity pulses. No patient required a blood transfusion.
At 24 h, color Doppler flow mapping revealed complete closure in all 43 patients (100%). No obstruction of the left pulmonary artery or the descending aorta was noted. Follow-up data were available in all 43 patients at one and three months after the procedure. All patients had complete closure with no evidence of device recanalization, migration, wire fracture, thromboembolism, endocarditis or hemolysis.
Transcatheter closure of PDAs is an established technique with no reported mortality and low morbidity. However, treatment in infants and closure of large PDAs is still associated with technical problems and a relatively high incidence of residual shunting (2–13). Such residual shunts have to be considered as failures because the patients remain at risk of bacterial endocarditis.
In the present study, transcatheter closure with the ADO was carried out in 43 patients with PDAs ranging from 2.2 to 8 mm at the smallest diameter. All PDAs were completely closed at 24 h with no complications during the procedure or at short-term follow-up. It appears, therefore, that the very high success rate and safety of the technique is related to the simplicity of deployment and the novel design of the device.
Comparison with the other devices
The ADO was designed to improve the closure rate of moderate-to-large sized PDAs regardless of anatomic configuration. Some of the previously used devices are the Rashkind double umbrella (C.R. Bard, Galway, Ireland) (2–4), the Sideris “buttoned” device (Custom Medical Devices, Amarillo, Texas) (5,6), the Botallooccluder (7)and the Gianturco spring coils (Cook Europe A/S, Bjaeverskov, Denmark) (8–13). Besides the coils, which can be delivered through a 4F sheath, all other devices have to be delivered through a 7F to 11F sheath, which makes their application difficult or impossible in infants and small children. The ADO can be delivered through a 5F to 6F sheath, which allows its use even in infancy. Moreover, in contrast to other occluders, the ADO stents the PDA, forcing the blood through a channel filled with highly thrombogenic polyester material, which should result in a virtually 100% occlusion rate by thrombosis. This obviates the need for multiple device implantation and reduces the cost of the procedure, the risk for embolization with tedious device retrievals or other complications. The previously reported incidence of residual shunting (3% to 38%) (2–13)is much higher than in the present study (0%), including those of Masura et al. (4%) (14). This excellent result in our study was obtained by significantly oversizing the devices more than recommended by the manufacturer (2 mm). The only drawback is the possibility of significant hemodynamic aortic obstruction by the retention disc.
An important advantage of the ADO is that it can be easily retracted into the delivery sheath and redeployed several times. This does not only reduce the risk for surgical or catheter removal of embolized devices but also reduces the cost obviating the introduction of a new device. With the exception of the detachable (controlled release) coils, all other devices are not repositionable, and once deployed, it is very laborious to reposition or to remove them. Misplacement of the ADO across the PDA occurred in one of our patients and was easily corrected by retracting the device into a large sheath and redeploying it.
As compared with other devices, the ADO’s implantation is much simpler without complicated mechanisms. This significantly reduces the fluoroscopy time, and shortens the learning curve for each operator. Indeed, the fluoroscopy time in this study was much lower (7.9 min [range 4.6 to 18 min]) than reported for simple (10,12)(14.8 to 40 min [range 4.2 to 152 min]) or detachable (13)(18.3 min [range 5 to 45 min]) spring coils, a technique that has recently gained wide acceptance as the treatment of choice (11,13).
The most significant complication of transcatheter closure of PDA has been reported to be device embolization to the pulmonary artery (2–13), and occasionally to the descending aorta (12). The previously reported embolization rate ranged from 3% to 20% (2–13). In the present study, the embolization rate was 0%. This is most likely due to the simple delivery mechanism, the design of ADO and the aforementioned oversizing of the devices.
Limitation of the study
In contrast to the preliminary work of Masura et al. (14), in this study, we successfully closed PDAs up to 8 mm in diameter of all angiographic types. However, the use of ADO, as well as the other currently available occluders, cannot be recommended for infants with moderate- to large-sized PDAs, in which surgical closure remains the treatment of choice. This is due to the fact that the device may protrude too much into the descending aorta or the pulmonary artery, causing significant obstruction. Because the PDA is a remnant of the sixth aortic arch, it forms an about 30° angle with the aorta. The retention disc of the ADO is at a right angle and, consequently, it will extend partially into the aorta, particularly with the B-type PDAs. In all the patients, this is not of hemodynamic significance, but in babies, it may cause partial aortic obstruction. A modification of the retention disc is in process, which may eliminate this drawback of the present device (Kurt Amplatz, personal communication). Before the release of the device, aortic pressure measurements should be carried out routinely. Protrusion into the aorta with a significant gradient requires removal of the device. However, protrusion into the left pulmonary artery results mainly in a redistribution of blood flow with no significant gradient. With growth of the child, the pulmonary arteries enlarge considerably, and this type of partial obstruction is expected to disappear at adolescence or long before, without clinical consequences (16). Further clinical testing is required to determine any potential limitations for the use of the ADO in patients with small and especially very small PDAs.
Our preliminary results prove that ADO is an efficient and safe device for transcatheter closure of PDAs. Further studies are required to establish long-term results in a larger patient population.
- Amplatzer duct occluder
- confidence interval
- patent ductus arteriosus
- pulmonary/systemic flow ratio
- Received June 4, 1999.
- Revision received October 15, 1999.
- Accepted November 19, 1999.
- American College of Cardiology
- Rashkind W.J,
- Mullins C.E,
- Helenbrand W.E,
- Tait M.A
- Hosking M.C.K,
- Benson L.N,
- Musewe N,
- Dyck J.D,
- Freedom R.M
- Rao P.S,
- Sideris E.B,
- Haddad J,
- et al.
- Rao P.S,
- Kim S.H,
- Cho J.Y,
- et al.
- Verin V.E,
- Saveliev V.S,
- Kolody S.M,
- Prokubovski V.I
- Moore J.W,
- George L,
- Kirkpatrick S.E,
- et al.
- Hijazi Z.M,
- Geggel R.L
- Hijazi Z.M,
- Geggel R.L
- Uzun O,
- Hancock S,
- Parsons J.M,
- Dickinson D.F,
- Gibbs J.L
- Galal O,
- de Moor M,
- Fadley F,
- et al.
- Masura J,
- Kevin P,
- Thanopoulos B,
- et al.
- Krichenko A,
- Benson L.N,
- Burrows P,
- et al.