|[Index FAC] [FVCC Index]|
Management of Patent ductus Arteriosus with
Emphasis on Transcatheter Therapy
Rao P. Syamasundar
Professor of Pediatrics and Medicine, Director Division of Pediatric Cardiology University of Texas-Houston Medical School.
In the fetus the ductus arteriosus, one of the fetal circulatory pathways, diverts the de-saturated blood from the main pulmonary artery into the descending aorta and placenta for oxygenation. After the infant is born, the ductus arteriosus constricts and closes spontaneously, presumably secondary to increased PO2. But in some infants, such spontaneous closure does not occur. This is more frequent in prematurely born infants than in full-term babies. Patent ductus arteriosus (PDA) may be an isolated lesion and may be present in association with other defects. Isolated PDA constitutes 6 to 11% of all CHDs . PDA is a muscular structure connecting the main pulmonary artery (at its junction with the left pulmonary artery) with the descending aorta at the level of left subclavian artery. The configuration of PDA varies considerably but most often it has a conical or funnel shape. The aortic end is wide and gradually narrows (ampulla) towards the pulmonary end. The narrowest segment is at the pulmonary end. Other types which are short and tubular and those with multiple constrictions and bizarre configuration can also be seen . Because of usually higher pressure and resistance in the systemic circuit than in the pulmonary circuit, left-to-right shunt takes place across the PDA. The magnitude of left-to-right shunting depends upon the diameter of the ductus and ratio of pulmonary to systemic vascular resistance. This presentation is limited to the discussion of isolated PDA beyond neonatal (and premature) period. In this review, I will present the treatment strategies to manage PDA with particular attention to transcatheter occlusion.
Since the description of successful surgical ligation of patent ductus arteriosus by Gross and Hubbard  in 1939, surgery has been extensively used in the treatment of PDA. Although surgical treatment has been shown to be safe and effective with only occasional complications, cardiologists have been attempting to develop less invasive, transcatheter methods of closure of PDA. The initial efforts of Porstmann [4,5] and Rashkind  and their associates paved the way to the development of a number of other PDA closure devices, reviewed elsewhere. [7-9] Most of these devices have been tested in animal models followed by human application. Design, redesign, testing and retesting and further refinement have been the case with most devices. The devices which did not reach the stage of human clinical trials are listed elsewhere  and will not be discussed. The devices thus far described and used in human clinical trials are listed in Table I. The selection of one device over the other is difficult because of lack of prospective randomized clinical trials. [10,11] Gianturco coils, Gianturco-Grifka Vascular Occlusion Device (GGVOD) and Amplatzer duct occluder are currently available for routine clinical use. The remaining devices, at the time of this writing, have not received approval from the appropriate regulatory authority (particularly FDA), but are available for use only at institutions which are participating in clinical trials with a particular device and will not be discussed.
Method of Implantation
In this section, the method of implantation of free Gianturco coils and Amplatzer Duct Occluder will be described. The delivery of detachable coils and GGVOD has been described elsewhere [12,13] and will not be reviewed here.
Gianturco Coils (0.038-in). The implantation of free coils was initially described by Cambier et al  and the method that we use [7,15,16] is similar that detailed by Cambier. These coils were originally described in 1975  and were used to occlude renal arteries and have undergone some changes over the years. They are now commercially available for clinical use and are made up of stainless steel wire with thrombogenic Dacron fibers attached to them. They are available in a variety of wire diameters, helical decimeters and lengths; we utilize 0.038-in diameter wire coils for PDA closure.
The procedure is performed with the patients sedated with a mixture of meperidine, promethazine and chlorpromazine, supplemented with intermittent doses of midazolam to ensure adequate sedation. Complete cardiac catheterization percutaneously via femoral vein and artery is performed to confirm the clinical and echocardiographic diagnosis and to exclude other cardiac defects. Heparin (75 to 100 units/kg) is given intravenously soon after the insertion of the arterial sheath. Aortic arch angiography in 300 right anterior oblique (RAO) and straight lateral views is performed with a 4- or 5-F pigtail catheter introduced through the femoral arterial sheath; 1 ml/kg of contrast material is used. Measurement of narrowest ductal diameter (usually at the pulmonary end), size of ampulla (at the aortic end) and length of ductus are measured in both views and averaged. These measurements are used as a guide to the selection of the diameter of the coil used for occlusion.
A 4-F right coronary artery catheter (Cordis, Miami , FL ) or a 4-F Glidecath catheter (Meditech, Watertown, MA) is introduced from the descending aorta into the main pulmonary artery via the PDA. If the catheter could not be advanced easily into the ductus, the soft end of a 0.035-in Teflon-coated (Cook), Benston (Cordis, Miami, FL) or angled floppy (Meditech) guide wire is used to cross the ductus. Once the tip of the catheter is positioned in the main pulmonary artery, this is confirmed by pressure measurements and if necessary, test injection of contrast material. Aortic arch angiographic frames, obtained in the RAO and straight lateral views, are used as a reference/guide throughout the procedure.
A coil with a loop that is 1.5 to 2 times the size of the narrowest diameter of the PDA is selected for implantation. The coil is loaded into the catheter with the stiff end of a 0.038-in Teflon-coated guide wire but is advanced with its floppy end. Under fluoroscopic monitoring (lateral view) one to one and one-half loops of the coil are delivered into the main pulmonary artery. The delivery guide wire is partially pulled back and the entire system (the coil and catheter) is withdrawn so that the delivered coil loops are drawn into the pulmonary end of the ductus. Then, the delivery catheter is pulled back into the aortic end of the ductal ampulla. The guide wire is re-advanced until it touches the coil in the catheter. The guide wire is fixed in position and the catheter is slowly withdrawn over the wire into the descending aorta, thus extruding the remaining coil into the aortic end of the ductal ampulla. Thus, the delivered coil straddles the narrowest diameter of the ductus. Fifteen minutes after coil implantation, repeat aortic arch angiography, careful pressure pullback from the aortic arch to descending aorta, and measurement of oxygen saturations from the superior vena cava, main pulmonary artery and descending aorta are performed. Three doses of Cefazolin (25 mg/kg/dose) are administered intravenously, the first dose is given in the catheterization laboratory and the second and third doses are given 6 and 12 hours after the first dose. The heparin is not continued nor its effect reversed. Clinical evaluation, chest roentgenogram and echo-Doppler studies are obtained on the day following the procedure and 1, 6 and 12 months after coil implantation.
Additional comments on the technique. Delivery catheter. Because of potential adverse effects of using larger sized catheters, we prefer to use 4-F catheters during retrograde delivery of the coil; we use either 4-F right coronary artery catheter (Cordis, Miami, FL) or 4-F Glidecath catheter (Meditech, Watertown, MA), both of which are suitable for implantation of 0.038-in coils.
Coil loop diameter. Because of potential for transient enlargement of minimal ductal diameters, especially in infants and younger children, we tend to use coils with large loop diameters (2 to 3 times the minimal ductal diameter), provided these diameters are smaller than ductal ampulla.
Number of coil loops. To provide better thrombus formation in the ductus  and to prevent re-canalization during follow-up,  we use 5-loop coils. However, if the ductal ampulla is shallow and unlikely to accommodate the longer coil, then, a 4-loop coil is utilized.
Modifications/refinements of the technique. Since the initial description by Cambier et al,  a number of modifications of the procedure have been suggested and include detachable coils, [12,18] antegrade and multiple coil techniques,  snare assisted coil delivery,  temporary balloon occlusion of the ductus on the aortic  or pulmonary artery  side, bioptome assisted coil delivery,  coil delivery via a tapered tip catheter,  double-disk shaped Diablo configuration design,  five loop coil design [15,16] and increasing the wire diameter to 0.052-in.  Some of the techniques have advantages over the conventional retrograde coil delivery, whereas others may not improve upon the procedure. Many of these changes increase the complexity of the procedure, prolong the fluoroscopy and procedure time or may increase the cost. These considerations should be taken into account when embarking on the use of modified techniques. Our own preference is to utilize conventional retrograde delivery of free 0.038-in Gianturco coils for very small PDAs (<1.5mm) and use 0.052-in coil, either by anterograde or retrograde fashion via 4-F long blue Cook sheath (Cook, Bloomington, IN) with the assistance of a bioptome for patients with minimal ductal diameters between 1.5 to 3 mm.  Larger PDAs (>3 mm) may require devices as detailed in a latter section of this paper.
Gianturco Coil (0.052-in). The procedure, up to the measurement of angiographic size of the ductus, is similar to that described in the second paragraph of 0.038-in coil implantation. The 052-in coils are similar to 0.038-in coils except that the wire diameter is 0.052-in. These coils can't be delivered via conventional catheters and therefore large bore catheters or sheaths are required. To prevent inadvertent delivery, bioptome assisted delivery is recommended.
The method that we use is similar to that described by Hays  and Grifka  and their associates. A coil with a loop that is 2 to 3 times the size of the narrowest diameter of the PDA is selected for implantation. Initially, the pigtail catheter, used for angiography, is removed and replaced with a 4-F right coronary artery catheter. This, with the help of a soft-tipped guide wire is advanced from the descending aorta, across the ductus into the main pulmonary artery. Following this, the guide wire is removed. A 0.25-in J-shaped guide wire is introduced through this catheter into the pulmonary artery and from there into the right ventricular apex. The guide wire is left in place, and the catheter removed. This is followed by removal of the short sheath and advancing a 4-F, long, blue Cook sheath with a radio-opaque marker at its tip. The dilator of the sheath and the guide wire are removed, and the tip of the sheath is left in place in the main pulmonary artery. The position of the sheath is confirmed by fluoroscopy and by pressure measurements. At this point, a 3-F bioptome is advanced through a coil loader (Cook). The bioptome tips are opened to capture the previously separated ball of 0.52-in coil. The coil is then withdrawn into the coil loader. The coil loader is then inserted into the hub of the 4-F long sheath and advanced within the sheath with the help of the bioptome until the coil reached the tip of the sheath. One loop of the coil is delivered into the main pulmonary artery, the entire system is withdrawn until the delivered coil loop engages into the pulmonary end of the ductus, and then the tip of the sheath is withdrawn into the descending aorta. The remaining loops are delivered into the ampulla of the ductus arteriosus by slowly advancing the bioptome. Having been assured of good position of the coil, the bioptome is opened, thus releasing the coil. If there is uncertainty of the coil position, it can be recaptured (withdrawn) into the sheath and the procedure repeated Fifteen minutes following coil occlusion, evaluation of the effectiveness of coil implantation is undertaken, similar to that described in the preceding section.
Amplatzer Duct Occluder. Musura and his associates  were the first to report clinical use Amplatzer duct occluder and the method that we use is similar that detailed by Musura. The device is constructed with 0.004-in thick Nitinol wire mesh, mushroom in shape and self-expandable in design. The devices are 5 to 8 mm long; the aortic end is 1 to 2 mm larger than the pulmonary end. A thin retention disc is located on the aortic side, 4 to 6 mm larger than the aortic diameter of the device. A recessed screw is built into the pulmonary end for connection to the delivery wire. Polyester fabric is sewn into the device to induce thrombosis after implantation. The available device sizes are 5/4, 6/4, 8/6, 10/8, 12/10, 14/12, and 16/14; the numbers indicate the diameter of the aortic and pulmonary ends, respectively. The devices, depending on the size of the device can be delivered transvenously via 5 to 7-F sheaths.
We implant the Amplatzer ductal devices also under conscious sedation and procedure is similar to coil occlusion up to the measurement of ductal size, as described in the second paragraph of coil occlusion procedure section. A 4 or 5-F multipurpose catheter is introduced into femoral vein, positioned in the main pulmonary artery and advanced into the descending aorta across the ductus. If the catheter can't be advanced across the ductus by itself, we use 0.035-in straight Benston guide wire, with a long floppy tip to cross the ductus. A 035-in extra-stiff exchange-length J-tipped Amplatzer guide wire is positioned in the descending aorta and the multipurpose catheter removed. (In rare occasions when ante-grade catheterization of PDA is not feasible, a guide wire (exchange length) is advanced into the pulmonary artery via a catheter introduced into the ductus from the descending aorta and the wire is snared from the pulmonary artery and exteriorized via the femoral venous sheath.). Then, an appropriate-sized Amplatzer delivery sheath is advanced over the wire, across the right heart and ductus and its tip positioned in the descending aorta.
A device whose pulmonary end is 1 to 2 mm larger than the size of the narrowest diameter of the PDA is selected for implantation. The selected Amplatzer device is de-aerated, screwed onto the loading wire and withdrawn, under water, into the loader sheath. After completely screwing the device, the device is unscrewed by one to one and one-half turns to facilitate unscrewing and release after it is implanted. The device is deposited into the delivery sheath already in place while flushing the device loader to avoid air entry into the system. The device is advanced within the sheath under fluoroscopic guidance. When the tip of the device reaches the tip of the sheath, the entire system is withdrawn until the tip of the sheath is in the descending aorta just distal to the aortic ampulla of the ductus. Then the sheath is retracted so as to uncover and deploy the aortic disc of the device. The entire system is slowly pulled back into the ductal ampulla and mid-ductus. An aortogram is performed to evaluate the position of the device. If satisfactory, the sheath is further withdrawn, while holding the device in place, to uncover the remaining part of the device in the ductus, across the narrow pulmonary end of the ductus. An aortogram is repeated to verify the position of the aortic disk within the ductus without protruding into the descending aorta and the position of the pulmonary end of the device across the narrowest part of the ductus and that there is no residual shunt around the device. Having been assured of good position of the device, the device is released; the delivery wire is rotated counterclockwise until releasing the device. The delivery wire and sheath are withdrawn into the inferior vena cava and the delivery wire is then taken out of the patient and the delivery sheath flushed. Ten minutes following device implantation/release, aortic arch cineangiography is performed in 300 RAO and straight lateral projections. Measurement of the pressures on pullback across the descending aorta and right ventricular, pulmonary arterial and aortic oxygen saturations are obtained. The follow-up is similar to that described for the coil occlusion section.
When feasibility, safety and effectiveness of most of the described techniques are carefully compared [11,12] they are found to be reasonably similar, although there are no carefully conducted randomized clinical trials. I will review the results of the three approved (FDA) methods mentioned above. [11,12] Residual shunts  hours after the procedure were present in 18% patients with free Gianturco coils which decreased to 9% at follow-up. With detachable coils, residual shunts were present in 7% to 28% immediately after the procedure which decreased further at follow-up to 3% to 12%. Residual shunts were present in 9% patients with GGVOD, all spontaneously closed during follow-up. Following implantation of the Amplatzer device, residual shunt were seen in 5% to 34% which decreased to 0 to 3% at follow-up.  In a multi-center US trial;  the Amplatzer device was successfully implanted in 435 (99%) of 439 patients. Complete occlusion by angiography was demonstrated in 384 (76%), which increased to 89% patients by echocardiography the following day. At one year follow-up, 359 (99.7%) of 360 patients had complete closure. Of the 35 consecutive Amplatzer ductal occlusions during a 25-month period ending August 2005, performed by the author, 34 had complete closure at implantation and the 35 th patient had complete closure at 3 month follow-up. At one-year follow-up, no residual shunts were seen.
The complications are minimal although coil/device embolization may occur, requiring transcatheter or occasionally surgical retrieval [10,11].
Amplatzer duct occluder has been used in closing moderate to large PDAs. [11,29-32] While it has been generally effective, its use in infants and young children has been reported to produce aortic obstrucion. [33,34] To circumvent this problem, the device was modified:  the retention disc was made thinner and concave and was build at a  degree angulation with the cylindrical long axis of the device such that the device conforms to the descending thoracic aorta. A platinum marker is placed in the downstream rim of the retention disc. Experimental evaluation  suggested that the objective of preventing aortic obstruction is achieved. Further modifications using finer wire mesh and without polyester disc was undertaken and successfully used in a single case.  Additional experience with this and other modified (swivel disc or disc-less) [39-41] devices appears to address the aortic obstruction in infants associated with the use of the conventional Amplatzer duct occluder. Larger experience with longer duration of follow–up is indicated.
Indications for Closure
Patients with clinical and echo-Doppler evidence of PDA who are ordinarily candidates for surgical closure are candidates for transcatheter occlusion. The procedure is indicated only in patients with continuous murmur suggestive of PDA with echo-Doppler confirmation. Closure is not recommended in the so called "silent ductus" detected incidentally without typical auscultatory features. Very small and small PDAs without hemodynamic overload are candidates for closure because of the risk of subacute bacterial endocarditis. Medium- and large-sized ductus should be closed to prevent further volume overloading of the left ventricle and to prevent pulmonary vascular obstructive disease apart from the endocarditis risk. Patients with persistent shunt following previous device occlusion should also have the residual ductus closed, mainly to prevent endocarditis. Similar recommendation may also be made for residual shunts following prior surgical ligation. Evidence for acute infection at the time of the procedure is a contraindication because of potential for development of bacterial endocarditis secondary to the foreign body (device). Following adequate control of infection, there is no contraindication for closure. There is no direct evidence to suggest that the procedure should not be performed in patients with coagulation or thromboembolic problems although, the effect of heparin, which is usually given during the procedure, should be considered. Patients with ductal dependent congenital cardiac anomalies and those associated with pulmonary vascular obstructive disease should not undergo closure.
Approaches to PDA Closure
Based on this and other reviews [7-11,15,27,40,41] and personal experience with PDA closure with a number of methods, I believe that the choice of the method of closure should be based on the shape  (Table II) and the size (minimal ductal diameter) of the PDA (Table III).
Ductal shape. Earlier descriptions such as conical, tubular, short and long have largely been replaced by classification described by Kriechenko.  If the ductus is very small or small, the shape may not play an independent role in determining the feasibility or effectiveness of closure. However, if the ductus is moderate-to-large, the shape may have special role on feasibility and effectiveness of device closure. In Table II are listed various ductal shapes and the devices that are likely to be useful in closing such ducts.
Size of the PDA. Silent PDA. Widespread use of color Doppler echocardiographic studies has resulted in identification of a new group of patients, commonly termed ‘silent ductus”. Nearly 1% of children undergoing color Doppler study were found to have a PDA.  These patients without auscultatory (continuous murmur) findings of ductus, but with visualization by color Doppler fall into the category of silent PDA. Some investigators [43,44] recommend closure while others  including our group [7,8,27] are opposed to closing such PDAs. The controversy continues but, as more data becomes available, the weight of evidence should eventually be the deciding factor.
Silent ductus after device/coil occlusion. Residual shunts may be present in a small number of patients following device/coil closure as reviewed in the preceding section. Most of these have no associated murmur and are technically “silent PDAs”. Can the non-interventional strategy be extended to such PDAs? This issue was examined by Latson et al  in a piglet models; the PDAs were occluded by Rashkind devices. Their studies suggested that the piglets with no residual and “trivial” shunts were not at a higher risk than controls whereas those with significant residual shunts were found to be at risk for developing endocarditis. Based on these data, Latson concluded that silent ductus after device occlusion are not at risk for developing endocarditis. But, it is not clear how these residual shunts, trivial vs. significant, compare with residual shunts, without murmurs in human subjects following device/coil closure. In addition, the current practice is to recommend antibiotic prophylaxis for prevention of subacute bacterial endocarditis if residual shunt exists after device/coil occlusion. Therefore, it appears prudent to recommend closure (mostly coil occlusion) if residual shunts are present beyond 6-12 months after initial attempt to occlude by device/coil.
Very small PDA. Our definition of very small PDA (Table III) is minimal ductal diameter < 1.5 mm with a continuous murmur. These PDAs can easily be occluded with free Gianturco coils, initially advocated by Cambier et al.  We prefer 0.038-in coils with 4 to 5 loops to effect complete occlusion. [15,16] These coils can be delivered via 4-F catheters, trans-arterially. The method is relatively simple and inexpensive. Use of snares,  forceps, [23,28] detachable coils [12,18] and balloon [21,22] and tapered tip  catheters during coil deployment is not necessary and Are likely to increase the cost and fluoroscopic time.
Small PDA. There are PDAs with minimal ductal diameters of 1.5 mm to 3 mm, again with a continuous murmur. Single 0.038-in Gianturco coil is likely to result in significant residual shunt and multiple coils may be required.  While multi-coil approach is a reasonable approach, because of potential for coil embolization and left-pulmonary artery stenosis, we do not advocate such an approach. For these patients, we employ 0.052-in coil;  these coils are delivered via long 4-F Blue Cook sheaths (Cook, Bloomington, IN) with the assistance of a biopsy forceps, a method similar to that described by Hays  and Grifka .
Whereas devices (Table III) may be used to occlude very small and small PDAs, the devices, in my opinion, are not necessary and may be problematic because of larger sheaths required to implant the devices and higher cost.
Moderate to large PDAs. These ducts with minimal ductal diameter > 3 mm, require closure and such closure may be accomplished by conventional surgical closure,  video-assisted throracoscopic interruption  and by a number of devices. The Rashkind PDA occluder, Botallo occluder, Clamshell device, polyvinyl alcohol foam plug and folding plug buttoned device have been discontinued or not available for routine clinical use. The usefulness of duct-occlude pfm  to close moderate to large PDAs has not been thoroughly investigated. Gianturco-Grifka sac  is available for general clinical use and is particularly useful in occluding tubular PDAs. However, a relatively large delivery sheath and difficulties in retrieving the dislodged devices are disadvantages with this device. Amplatzer duct occluder [11,29-32] has several favorable features, including relatively small delivery sheath (5 to 7-F), facility with which it can be retrieved and repositioned prior to detachment and high closure rates. However, device dislodgment requiring surgery [31,32] is of some concern although the device dislodgment rate (3 of 209 or 1.4%) in large series  is low. Finally, wireless devices, such as transcatheter patch  may be useful in occluding large PDAs once they become available for clinical use. Thus, a number of devices [9-11] have been used to occlude moderate to large PDAs; some have been discontinued and others modified. Some are available for general clinical use and others are under clinical trial protocols. At the current stage of development of device technology, Gianturco Grifka device and Amplatzer duct occluder compete with conventional surgical closure and video-thoracoscopic closure. The selection of the method for occlusion of moderate to large PDA would largely depend upon the availability of a particular device or method at a given institution at a given time. When several devices become available for general clinical use, an opportunity to perform randomized clinical trials may arise and such data may help device on the best method of PDA closure.
Summary and Conclusions
Because of availability of several devices, the selection of devices for implantation is now largely dependent on angiographic minimal ductal diameter measured immediately preceding the placement of closure devices. [11,41] We do not recommend closure of silent PDAs.  Very small to small PDAs can be successfully closed with the free or detachable Gianturco coils. Small to moderate PDAs may need multiple coils  or larger wire diameter (0.52-in) coils. [23,26-41] Options for closure of moderate-to-large PDAs are devices, video-thoracoscopic interruption and conventional surgical closure. Among the devices, GGVOD is approved for general clinical use and is useful in tubular PDAs. Amplatzer Duct Occluder has been approved by the FDA and is useful in closing moderate to large PDAs. [11,29-32] Duct-occlud pfm is currently undergoing FDA-approved clinical trials and is available for investigational use at the participating hospitals. Wireless PDA devices  may be used for very large PDA and are undergoing clinical trials outside USA . Availability of the method or device and expertise at a given institution at a given time are likely to determine the method selected.
- Professor of Pediatrics and Medicine
Publicación: November 2005
FORM DEACTIVATED SINCE
December 1st., 2005
Your questions, contributions and commentaries will be answered
by the lecturer or experts on the subject in the Pediatric Cardiology list.
Please fill in the form and press the "Send" button.
Dr. Diego Esandi
Dra. Silvia Nanfara
Prof. Dr. Armando Pacher
Webmaster Updated: 11/30/2005