Transcatheter Closure of Secundum Atrial Septal Defects with Complex Anatomy (Part I)

ABSTRACT: The aim of this study was to evaluate the feasibility, safety and efficacy of transcatheter closure of secundum atrial septal defects (ASD) in patients with complex anatomy. From September 1997 to July 2003, a total of 40 patients (median age, 34 years; 65% female) with complex ASDs, defined as the presence of a large defect (stretched diameter > 26 mm) associated with a deficient rim (n = 23); multiple defects (n = 8); a multi-fenestrated septum (n = 5); and defects associated with an aneurysmal septum irrespective of their size (n = 4) underwent closure. The Helex device was used in 4 patients and the Amplatzer in the remaining. Two devices were implanted in 2 patients each. Implantation was unsuccessful in 5 patients, with 4 having large defects associated with a deficient anterior rim and a floppy posterior septum. Occlusion was observed in 22/35 patients (63%) immediately after implantation and in 31 (89%) at a mean follow-up of 18 ± 9 months. No major complications occurred. Right ventricular end-diastolic dimensions (indexed for body surface area) decreased from 135 ± 25% before closure to 124 ± 15% 24 hours after closure, and to 92 ± 12% after 12 months. Two patients with 2 distant defects and 2 patients with large defects remained with shunts (Key words: atrial septal defect, septal occluder device Transcatheter closure of secundum atrial septal defects (ASDs) has evolved over the last 2–3 decades since its first description by King and Mills in 1976.^1–10 Classically, selection criteria for adequate device placement included a single and centrally located defect with sufficient (> 4–5 mm) and firm surrounding rims. However, anatomical suitability for transcatheter closure has also evolved due to the versatile designs of some devices, especially the Amplatzer septal occluder (ASO; AGA Medical Corporation, Golden Valley, Minnesota), and refinements in implantation techniques. The aim of this study was to evaluate the feasibility, safety and efficacy of transcatheter closure of secundum ASDs with complex anatomy. Methods Definitions. For transcatheter purposes, secundum ASD with complex anatomy was arbitrarily defined as the presence of a large (stretched diameter > 26 mm) ASD associated with a deficient ( 10 mm), considered to be aneurysmal, were also regarded as ASDs with complex anatomy, irrespective of their size. Study subjects. From September 1997 to July 2003, a total of 143 patients with hemodynamically significant secundum ASDs with increased right ventricular end-diastolic dimensions (indexed for body surface area) and abnormalities in ventricular septal motion by transthoracic echocardiography (TTE) underwent an attempt of transcatheter closure mainly at our institutions (n = 104), but also at other centers in our country (n = 39). All attempts were performed or supervised by one of the operators (CACP, CAE, SLNB, VFF). Selection of patients with regard to transcatheter closure suitability was carried out by ambulatory transesophageal echocardiography (TEE) days to months prior to the procedure. Out of the 143 patients, forty (28%) were considered to have complex anatomy as defined above. Two patients had significant associated lesions. A 9-year-old girl had a 2.5 mm patent ductus arteriosus, which was occluded in the same catheterization session with a coil.¹¹ A 51-year-old patient with a large defect (26 mm on TEE) had impaired left ventricular function (ejection fraction, 38%) due to an old anterior infarct. He was admitted to the hospital 24 hours before the procedure to receive intravenous diuretics and dobutamine in order to improve left ventricular diastolic function and prevent pulmonary edema after ASD closure.^12,13 Patients with complex ASD anatomy ranged in age from 5–60 years (median age, 34 years) and weighed from 18–90 kg (median weight, 65 kg). Twenty-six patients (65%) were female. Informed consent for the procedure was obtained from all patients or guardians. Devices and implantation procedure. The Helex device (William Gore and Associates, Flagstaff, Arizona) was used in 4 patients and the ASO device was used in all others. Device description and the basic implantation techniques for both devices have been previously reported.^5,6,9,10 Modifications in the implantation technique for the ASO were required in some patients and are described below. All procedures were performed under general endotracheal anesthesia and TEE guidance.^14,15 From September 1997 to November 1999, balloon sizing of the defect was performed using the dynamic method, as previously reported. A Meditech sizing balloon (Boston Scientific/Scimed, Inc., Maple Grove, Minnesota) was inflated with diluted contrast in the left atrium and brought toward the atrial septum until the defect was occluded, as documented by TEE. The balloon was gradually deflated, with gentle traction applied to the catheter. This was recorded on cine and TEE, and the diameter of the balloon at the moment it popped to the right atrium was considered to be the stretched diameter. A sizing plate was also employed to further assess the balloon diameter at that time. Since November 1999, the stationary method has been employed using a 24 or 34 mm sizing balloon (AGA Medical Corporation). The balloon was inflated with diluted contrast media until the appearance of a waist and disappearance of left-to-right shunting across the defect (“stop-flow”). This was recorded on cine and TEE, and the size of the waist, which was considered to be the stretched diameter, was measured by both methods. In some patients, a second and smaller defect was only demonstrated after balloon occlusion of the larger one. When present, the size and location of the smaller defect was recorded, as well as the distance between both defects. When the defects were close to each other ( 7–8 mm), both were crossed with different catheters and 2 sizing balloons were positioned across the atrial septum to estimate the stretched diameter of each hole. In patients with multi-fenestrated septums, the most centrally located defect was crossed with a catheter and the stretched diameter was determined in the usual way. Defects associated with an aneurysmal septum were stretched using the least amount of diluted contrast medium in the balloon (static method) needed to generate a waist. In general, ASO size was selected to be 0–2 mm larger than the stretched diameter. Oversizing the ASO up to 2–4 mm larger than the stretched diameter was employed in patients with large defects (stretched diameter > 26 mm) and a deficient (9,10,16 the size of the Helex device was chosen to be at least 80–100% larger than the stretched diameter (max, 20–22 mm), with care taken not to overstretch the defect with the sizing balloon. The choice between ASO or Helex device was based on the availability of the device and the underlying anatomy. The Helex device was introduced in our country only in March 2002. From then on, it was primarily employed in patients with a fenestrated septum and with a small to moderate ASD associated with an aneurysmal septum, such as seen in 2 patients in this series. In patients with 2 defects close to each other ( 7–8 mm), two separate devices (either ASO or Helex) were implanted in each defect according to previously published protocols.^11,16,17 The smaller defect was the first to have a device deployed and not released. This was followed by implantation of a second device in the larger defect. After satisfactory positioning of both devices was confirmed by TEE, release from the delivery cable was carried out starting with the first implanted device. The implantation technique of the ASO has evolved since the beginning of this experience and was modified according to the underlying anatomy and failure to implant the device at the first attempts using the standard technique. In patients with large defects and deficient anterior rims, the left atrial disc of the Amplatzer device would often prolapse through the defect during deployment of the connecting waist, coming in at a perpendicular angle toward the atrial septum. To overcome this problem, some maneuvers were employed according to operator preference. First, the left atrial disc was initially deployed in the mouth of the left upper pulmonary vein, assuming a “bubble” appearance. The sheath was retracted to deploy part of the device in the left atrium and part in the right atrium. While retracting the sheath, gentle traction was applied to the delivery cable. This generally enabled the device to “catch” the atrial septum after both discs were reconfigured. Second, the left atrial disc was initially deployed at the mouth of the right upper pulmonary vein, allowing proper alignment of the left atrial disc with the atrial septum. The latter option was also employed in a patient with borderline (6–7 mm) posterosuperior rim (toward the superior vena cava). Third, a slight curve (30º) was manually shaped onto the distal segment of the delivery cable so that during deployment of the waist and the right atrial disc, the left disc would come parallel to the atrial septum. However, by doing so, in order to unscrew the device, the long sheath had to be re-advanced to rectify the delivery cable. After the sheath was gently touching and securing the device, the delivery cable was rotated in a counterclockwise fashion to unscrew and release the device. Sometimes, depending on its availability, instead of remodeling the delivery cable, the long sheath was changed for one with a very sharp curve (Cook Cardiology, Bloomington, Indiana) to achieve the same angle to attack the atrial septum. We did not cut the tip of the long sheath in an oblique fashion to enable the device to exit at an angle from the delivery sheath. TEE monitoring. A complete TEE study was performed under general anesthesia just before the procedure. For complete assessment of the defect and the atrial septum, standard echocardiographic views such as four-chamber, short and long axis were employed. Generally, the superior and inferior rims were measured using the long-axis view, which demonstrated the superior and inferior vena cava in the same plane. The anterior and posterior rims were assessed using the short axis view at the level of the aortic root. Patients with a distance of 4 mm) according to the Toronto protocol.¹⁸ Post-procedure care and follow-up. Cephazolin (20 mg/kg; max, 1 gram) was given during the procedure and at 8-hour intervals (total, 3 doses). Hemostasis was achieved by manual compression. The patients were awakened in the catheterization aboratory and transferred to the recovery room for routine clinical observation. They were discharged home the following day and instructed to receive aspirin (2–5 mg/kg/day; max, 100 mg) for 6 months, to avoid contact sports for 2–3 months and to observe the recommendations for endocarditis prophylaxis for 6 months or until complete closure was documented. Chest radiograph, electrocardiogram and TTE were obtained before discharge. Right ventricular end-diastolic dimension measured by TTE was indexed for body surface area according to previously published protocols.^3,7 TTE and TEE were performed after 6–12 months to assess chamber sizes, ventricular septal motion, device position and potential residual shunting. In the presence of residual shunting, TEE was repeated after 12 months. Statistical analysis. Values are expressed as means and standard deviations or medians and ranges, as appropriate. Chi-square or Fisher’s exact tests were used to assess changes in interventricular septal motion. ANOVA tests were employed to assess changes in the right ventricular end-diastolic dimensions with time. The level of significance was set at p Procedure results. We were unable to place the device across the atrial septum in 5 patients, four of whom were in the first half of our experience (the first 70 patients). These 4 patients had large defects (stretched diameters: 26 mm, 28 mm, 30 mm and 32 mm) associated with a deficient anterior rim and a thin, floppy and hypermobile posterior rim (Figure 1). In these patients, the Amplatzer devices would not anchor in the atrial septum despite repeated trials using the standard and modified techniques. In 2 patients, the posterior septum was torn after repeated attempts at anchoring the device. The remaining patient had a 10 mm ASD by TEE with a stretched diameter of 13 mm associated with an aneurysmal and redundant atrial septum. Despite satisfactory positioning of a 25 mm Helex device, the device was not properly locked (missed one eyelet) before final release. After successful device removal, we discovered the thin septum was torn due to catheter and device manipulation and the procedure was abandoned. All 5 patients with failed implantation had uneventful recoveries and underwent successful surgical closure of the ASDs during the same hospital admission. Continued on next page