Transcatheter Closure of Small-to-Large Patent Ductus Arteriosus
with Different Devices: Queries and Challenges

Mehnaz Atiq, MD, Nadeem Aslam, MD, *Khawar A Kazmi, MD
Mehnaz Atiq, MD, Nadeem Aslam, MD, *Khawar A Kazmi, MD

Patent ductus arteriosus (PDA) can cause congestive cardiac failure, repeated pneumonia, pulmonary hypertension and an increased risk for endocarditis. Transcatheter closure of PDA is currently the preferred therapeutic alternative to surgical ligation1–3 in infants, children and adults. During the last decade, clinical experience has widened extensively and several occlusive devices have been used for this purpose. These include Rashkind’s double umbrella, the Sideris buttoned device, Gianturco spring coils, Cook’s detachable coils and, most recently, the Amplatzer duct occluder.3– 5 In the present era, surgical ligation is reserved for large symptomatic PDAs in very small infants and premature babies, unfavorable morphology of the duct5,6 or due to cost factors, since surgical ligation may be a cheaper alternative. We present ourexperience using transcatheter closure for small-to-large PDAs in infants, children and adults, and postoperative residual PDAs. In addition, we discuss the challenges we faced using this treatment approach.

Patients and Methods

Ninety-eight consecutive patients who underwent attempted transcatheter closure of PDA during a 6-year period (2001–2006) were included in this retrospective review. Clinical history and examination pertinent to PDA were noted. The mean age of patients was 64 ± 11 months (range, 7 months to 54 years). Thirty-seven patients were infants and 10 were adult patients. There were 66 females and 32 males. Two patients had residual PDA after surgical ligation. Sixty six patients were symptomatic with respiratory infections, congestive cardiac failure (or failure to thrive), and 32 were asymptomatic. The mean weight was 14.4 ± 1.6 kg (range 5–52 kg), with a mean in the 10th percentile for weight (according to age) in infants and children. Four children had trisomy 21, 3 had congenital rubella syndrome and 1 had acute lymphoblastic leukemia in remission. Additional associated cardiac diagnoses were seen in 16 patients (Table 1).

Chest X-ray and electrocardiogram findings were noted. Color flow mapping (2-D, M-mode) and spectral Doppler echocardiography were performed using Sonos 5500 Philips echocardiography equipment (Philips Medical Systems, Andover, Massachusetts), with a 5 MHz transducer for children, and 3.75 MHz for adult patients. Patients were examined in the parasternal long- and short-axis projections for left ventricular diastolic dimension and ratio of left atrium-to-aortic root diameter, high short-axis and suprasternal projections for sizing of the ductus arteriosus. Noninvasive estimate of systolic pulmonary arterial pressure and associated cardiac anomalies were noted.

Patients weighing more than 4 kg were selected. The procedure was performed under conscious sedation with oral chloral hydrate 75 mg/kg and intravenous nalbuphine 0.1–0.2 mg/kg. Adult patients did not receive any sedation. An initial aortogram in lateral projection (90º left anterior oblique), or its modification, was done to profile the size and shape of the PDA. Ducts measuring less than 2.5 mm in the narrowest diameter were closed with detachable coils, and those measuring more than 2.5 mm were occluded with the Amplatzer Duct Occluder (ADO). For large ducts with severe pulmonary hypertension (systolic pulmonary arterial pressures equal to systemic systolic pressures), the Amplatzer muscular VSD device was selected for occlusion after documenting reversibility with 100% oxygen inhalation. All patients received intravenous heparin 100 IU/kg after arterial access to achieve an activated clotting time of approximately 200 seconds, and intravenous cefazolin 40 mg/kg prior to device implantation.

Device description, selection and implantation techniques. Detachable coils (Cook Cardiology, Bloomington, Indiana) are made of stainless steel covered with thrombogenic Dacron wool fibers. The sizes available are 3 mm, 5 mm, 6.5 mm and 8 mm in diameter, with 3, 4 or 5 loops. The coil diameter selected was twice the diameter of the narrowest part of the PDA. The coil has a helical screw that is screwed on to the delivery system that consists of an outer hollow cable and an inner straightening mandrel to keep the coil straight during its advancement through a 5 Fr delivery catheter. Coils are usually deployed retrogradely via the femoral arterial approach. One to 1.5 loops of the coil are deployed in the pulmonary artery, 0.5 to 1 in the duct, and the remaining 2 to 3 loops in the aortic ampulla. The coil is then released after confirming its stability. The postprocedure aortogram was performed after 10 minutes, and if significant residual leakage was noted, a second coil was deployed. ADO (AGA Medical Corporation, Golden Valley, Minnesota) is a self-expanding, self-c enter ing, mushroom-shaped, low-profi le device made of nitinol wire mesh into which are sewn thrombogenic polyester fibers. It has a flat aortic r e t e ntion disc, 4–6 mm larger than the cylindrical body whose length is 7–8 mm. The device sizes range from 6/4 to 16/14 with 2 mm increments, and the size selected is 2–4 mm larger than the smallest diameter of the PDA. The Amplatzer muscular VSD device is similar to the ADO, except that it has a retention disc at both ends. Both of the devices have a screw at the pulmonary end that is attached to a cable through a loader. The device is pulled into the loader under water seal to minimize air trapping, and pushed through a transvenously-placed long sheath that crosses the PDA into the descending aorta. The aortic retention disc is deployed into the aorta and the whole device-delivery sheath assembly is then pulled back into the ampulla after which the cylindrical part is deployed, thereby plugging the PDA. An aortogram is performed prior to the release of the device.

A follow-up echocardiogram was performed 1 week later and if residual leakage was found, it was repeated after 1 month and 6 months if required. Echocardiography was done in all patients with pulmonary hypertension to evaluate the noninvasive estimates of pulmonary arterial pressures at 1 month and 6 months after the procedure.

Statistical analysis. Statistical analysis was conducted using SPSS version 13 (SPSS, Inc., Chicago, Illinois). Frequencies, means and standard deviations were calculated by descriptive statistics. Univariate analysis was performed to determine the differences between the coil and the ADO group. A p-value < 0.05 was considered significant.


The mean pulse pressure was 49 ± 5 mmHg (range 35–70 mmHg). A continuous murmur was audible in 68 patients, a systolic murmur was heard in 21, and 9 patients had no murmur. Chest X-ray showed cardiomegaly in 40 patients and pulmonary plethora in 17. The electrocardiogram was normal in 45 patients; left ventricular forces were seen in 35 patients, and biventricular forces were noted in 18. On echocardiography, the mean LA/AO ratio was 1.6 ± 0.2 (range 1.3–1.8), and the mean size of the PDA was 3.3 ± 1.1 mm (range 1.1–9 mm). Pulmonary hypertension was found noninvasively in 19 patients. Thirty patients required anticongestive therapy with furosemide, and 9 were prescribed captopril. Three were on sildenafil for severe pulmonary arterial hypertension.

Cardiac catheterization revealed a mean pulmonary arterial systolic pressure (PASP) of 48 ± 11 mmHg and a mean pulmonary arterial mean pressure of 33 ± 13 mmHg. Seventeen patients had pulmonary arterial systolic pressures > 40 mmHg, and moderately severe pulmonary hypertension was noted in 8 patients (PASP 69 ± 26 mmHg in this group). The mean ratio of systolic pulmonary arterial-to-systolic aortic pressure was 0.35 ± 0.11 (range 0.14–1.0). The mean size of the PDA on aortography was 3.1 ± 1.4 mm (range 1.1–11 mm). After the initial aortogram, 7 patients were considered unsuitable for device occlusion due to an unfavorable size or shape of the PDA and were referred for surgical ligation. Thirty-seven patients underwent closure with a detachable coil, 52 with an Amplatzer duct occluder, and 2 were closed with an Amplatzer muscular VSD device. The children who had a muscular VSD device implanted in the PDA had systolic pulmonary arterial pressures equal to systemic systolic pressures and had balloon occlusion prior to PDA closure. One patient required occlusion with 2 coils. The mean fluoroscopic time was 9.2 ± 5.5 minutes (range 3.4–26 minutes). The differences between patients implanted with a coil versus an ADO are shown in Table 1. In 5 patients, associated cardiac defects were also treated by interventional techniques. These were closure of ASD in 2, and balloon pulmonary valvuloplasty in 1.

Some PDAs were a challenge to address. In 1 patient with a large, long duct, the device had to be pulled inside the duct so as to prevent narrowing of aortic lumen. In a baby with Down’s syndrome, a small duct went into a spasm and would not allow passage of the coil delivery catheter. The duct negotiation was performed with the straightened coil, resulting in successful closure. There was embolization of an Amplatzer muscular VSD device to the right pulmonary artery in 1 patient with a large unfavorably-shaped PDA. This occurred 12 hours after implantation and was retrieved surgically. In another patient, an initial 5PDA5 coil was changed to a larger 6.5 PDA5 coil because of suspected instability on gentle pushing before deployment. Ten patients had loss of dorsalis pedis pulse after the procedure which returned within 6–18 hours of intravenous heparin infusion. Two patients who had an initial hematocrit of 24% and 26% required blood transfusion due to significant blood loss during the procedure and arterial access.

The mean occlusion rate with coils in the catheterization laboratory was 84 ± 7%, and was 83% with ADO. After 1 week, the mean occlusion rate with coils was 96%, and 99% with ADO. One patient with a large PDA requiring a 14/12 ADO was left with a mild gradient of 9 mmHg in the aorta at the end of the procedure. Patients were followed up for 2–16 months. None of our patients had stenosis of the left pulmonary artery and 1 has a residual gradient of 10 mmHg in descending on echocardiography. No patient had residual ductal patency after 10 months. In patients with pulmonary hypertension, there was normalization of noninvasive pulmonary pressure estimates in 1–6 months.


Patent ductus arteriosus constitutes 8% of congenital heart disease, being twice as common in females,1 as was also coincidentally seen in our study. The first attempt at transcatheter closure of PDA was made in 1967 by Portsman et al7,8 using an Ivalon plug. Thereafter, several devices were introduced, but most had either high embolization or residual shunt rates9,10 and were withdrawn from use. The devices currently in use include Gianturco coils, Gianturco-Grifka vascular occlusion devices, the Duct occluder PFM coil, the detachable Cook’s coil and the Amplatzer duct occluder.1 Detachable coils are more expensive than the nondetachable Gianturco coils, but have very infrequent embolization rates.3 The Amplatzer duct occluder has emerged as a very successful, less invasive alternative to surgery to close moderate-to-large ducts.10,11 Multiple coils could also be used to close moderatesized ducts, but have the disadvantages of occasional stenoses of the left pulmonary artery and long fluoroscopy times.12,13

In general, the advantages of the coil are its low cost, small sheath size and shorter fluoroscopic times. The advantages of ADO are that its deployment is simple, it is retrievable and repositionable and good occlusion rates can be achieved.4,5 However, the sizes of the delivery sheath, and the shapes and sizes of the available occluders limit its use in small infants and small ducts. In addition, in infants weighing less than 4 kg, the standard ADO may not be appropriate due to protrusion of the occluder either into the pulmonary artery or the aorta. The angled ADO14 was designed to address this issue of protrusion of the aortic disc. The aortic disc is angled at 32° to the body of the device and concave towards the aorta to prevent protrusion into the aorta. It appears promising in small infants and PDAs with a small ampulla.

The occasional problem of ductal spasm during aortography can lead to significant undersizing and device embolization.15 This, however, did not happen in our series of patients. Two of our patients had a residual PDA after an initial surgical ligation. Podnar et al16 report on 12 cases of residual PDA after surgery, which were later closed using Gianturco coils in 2 patients, detachable coils in 9 and an ADO in 1. In our 2 patients who had residual postoperative patency, the duct measuring 2 mm in diameter was closed with a coil, while the other duct measuring 5 mm was occluded with an ADO.

Hemolysis due to intradevice or residual leakage has been reported with both coils and ADO.17 Today, this problem is generally encountered less frequently. In our series, hemolysis was not seen in any patient. Large PDAs have been successfully closed with ADO.11,12 However, other devices have also been used to close large PDAs like the Amplatzer septal occluder and the Amplatzer muscular ventricular septal defect device.18,19 The largest PDA in our series measured 11 mm in a 5-year-old girl. This, along with 1 measuring 8 mm in diameter, which was seen in a 7-year-old patient, were closed with an Amplatzer muscular VSD device. Embolization was seen in the former.

Thirty-two of our patients were asymptomatic; 9 of these had no distinct clinical murmur. The management of this subset of patients who have a “silent duct” is controversial. Many recommend closure to prevent endocarditis, while others advise against implanting a foreign body and incurring the increased cost of patient management.20

Midterm follow up of our patients showed a complete occlusion rate of 99.5% for all the types of devices we used. Long-term encouraging results have also been reported by several authors worldwide.21–23


Transcatheter closure of PDA is the preferred alternative to surgical ligation. Complications are minimal, and newer devices have further made them a rarity. However, a small subset consisting of large unfavorably-shaped ducts will continue to be challenging, even to the most experienced operators.



1. Rao PS. Summary and comparison of patent ductus arteriosus closure methods. In: Rao PS, Kern MJ (eds). Catheter-based Devices for Treatment of Non-Coronary Cardiovascular Disease in Adults and Children, 1st Edition. Philadelphia: Lippincott Williams and Wilkins, 2003, pp. 219–228.

2. Lorber A, Gazit AZ, Khoury A, et al. Percutaneous transaortic occlusion of patent ductus arteriosus using a new versatile angiographic and delivery catheter. Pediatr Cardiol 2003;24:482–483.

3. Moore JW, Levi DS, Moore SD, et al. Interventional treatment of patent ductus arteriosus in 2004. Catheter Cardiovasc Interv 2005;64:91–101.

4. Bilkis AA, Alwi M, Hasri S, et al. The Amplatzer Duct Occluder: Experience in 209 patients. J Am Coll Cardiol 2001;37:258–261.

5. Sandhu SH, King TD. Amplatzer duct occluder. In: Rao PS, Kern MJ (eds). Catheter-based Devices for Treatment of Non-Coronary Cardiovascular Disease in Adults and Children, 1st Edition. Philadelphia: Lippincott Williams and Wilkins, 2003, pp. 213–218.

6. Rao PS. Coil occlusion of patent ductus arteriosus (editorial). J Invasive Cardiol 2001;13:36–38.

7. Porstmann W, Wierny L, Warnke H, et al. Catheter closure of patent ductus arteriosus: 62 cases t reated without thoracotomy. Radiol Clin North Am 1991:9:203–218.

8. Wierny L, Plass R, Porstmann W. Transluminal closure of patent ductus arteriosus: Long-term results of 208 cases treated without thoracotomy. Cardiovasc Interv Radiol 1986:9:279–285.

9. Report of European Registry. Transcatheter closure of persistent arterial duct. Lancet 1992;340:1062–1066.

10. Jan SL, Hwang YC, Chi CS. Transcatheter closure of a large patent ductus arteriosus in a young child using the Amplatzer duct occluder. Pediatr Cardiol 2005:26:703–706.

11. Jan SL, Hwang B. Fu YC, Chi CS. Transcatheter closure of a large patent ductus arteriosus in a young child using the Amplatzer Duct Occluder. Pediatr Cardiol 2005;26:703–706.

12. Santoro G, Bigazzi MC, Palladino MT, et al. Comparison of percutaneous closure of large patent ductus arteriosus by multiple coils versus the Amplatzer duct occluder device. Am J Cardiol 2004;94:252–255.

13. Podnar T, Gavora P, Masura J. Percutaneous closure of patent ductus arteriosus: Complementary use of detachable Cook patent ductus arteriosus coils and Amplatzer Duct Occluder. Eur J Pediatr 2000;159: 293–296.

14. Masura J, Gavora P, Podnar T. Transcatheter occlusion of patent ductus arteriosus using a new angled Amplatzer Duct Occluder: Initial clinical experience. Catheter Cardiovasc Interv 2003;58:261–267.

15. Ewert P. Challenges encountered during closure of patent ductus arteriosus. Pediatr Cardiol 2005;26:224–229.

16. Podnan T, Masura J. Transcatheter occlusion of residual patent ductus arteriosus after surgical ligation. Pediatr Cardiol 1999;20:126–130.

17. Uzun O, Dickinson D, Parsons J, et al. Residual and recurrent shunts after implementation of Cook detectable dual occlusion coils. Heart 1998;79:220–222.

18. Thanopoulos BD, Tsaousis GS, Djukic M, et al. Transcatheter closure of high pulmonary artery pressure persistent ductus arteriosus with Amplatzer muscular ventricular septal defect occluder. Heart 2002;87:260–263.

19. Santoro G, Begazzi MC, Carrozza M, et al. Percutaneous treatment of moderate to large patent ductus arteriosus with different devices: Early- and mid-term results. Ital Heart J 2005;6:396–400.

20. Balzer DT, Spray TL, McMullin D, et al. Endocarditis associated with a clinically silent patent ductus arteriosus. Am Heart J 1993;125:1192–1193.

21. Masura J, Tittle P, Gavora P, Podnar T. Long-term outcome of transcatheter patent ductus arteriosus closure using Amplatzer duct occluders. Am Heart J 2006;151:755.

22. 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.

23. Hosing MCK, Benson LN, Musewe N, et al. Transcatheter occlusion of persistently patient ductus arteriosus: Forty month follow up and prevalence of residual shunting. Circulation 1991;84:2313–2317.