Transcatheter and intraoperative device closure of atrial septal defect in infants under three years of age

This study was approved by the Ethics Committee of Linyi People’s Hospital Affiliated to Shandong University. An informed written consent was obtained from each patient for every procedure.

From September 2010 to September 2018, 186 children under 3 years of age with significant secundum ASD underwent ASD device closure. The patients were divided into two groups in accordance with the closure approach. A total of 88 and 98 patients were included in groups A (transcatheter device closure) and B (intraoperative device closure), respectively. The indication for ASD closure was hemodynamically significant left-to-right shunt (pulmonary to systemic flow ratio > 1.5 measured by echocardiography), which was manifested by the enlargement of the right ventricle (RV). Other indications included failure to thrive, frequent respiratory infections, increased sweating and easy fatigue, significantly elevated pulmonary arterial pressure, or strong parental request. All children were diagnosed with secundum ASD without the presence of other cardiac malformations, elevated nonreactive pulmonary vascular resistance, uncontrolled congestive heart failure, and any infection that could not be successfully treated before device closure. In general, the inclusion criteria for group A were as follows: secundum ASD with the presence of adequate rims (≥5 mm); the distance from the defect edge to the coronary sinus, superior and inferior vena cava, and pulmonary vein was ≥5 mm, and that until the atrioventricular flap reached 7 mm; suitable body weight(≥10 kg) with good peripheral vascular development. The inclusion criteria for group B were as follows: inclusion criteria for group A; no specific requirement for patient weight; secundum ASD with deficient rim < 1/4. Before treatment selection, all the guardians of the patients were informed of the indications, advantages and disadvantages, and specific risks of both treatments. For example, transcatheter device closure involves a short postoperative recovery time and zero incision. The intraoperative device closure requires minimal incision but no vascular injury and X-ray exposure and can be immediately converted to surgical repair in the operating room. In general percutaneous closure is preferred for patients who satisfied the inclusion criteria for group A, whereas the final treatment plan must be considered along with the preferences of the patient’s guardian. Table 1 summarizes the patients’ clinical data regarding sex, age, body weight, ASD sizes, indications, and characteristics. Routine clinical examinations before the procedure included transthoracic echocardiography (TTE), chest X-rays, electrocardiogram, and blood and biochemical tests. The ASD diameter and atrial septum length were measured in several views by echocardiography, and the largest measurement was recorded as the ASD diameter. Pulmonary pressures were assessed by the tricuspid and pulmonary regurgitations and ventricular septal shape.

A domestic ASD device (Shanghai Shape Memory Alloy Co., Ltd., Shanghai, China) made from nickel and titanium alloys was used in both groups. The ASD occluder was modified from the Amplatzer atrial septal occluder. A standard transfemoral approach was adopted in group A. The delivery system, which consisted of a sheath and a pushing rod, used in group B was shorter than that in group A, which was specially designed for intraoperative device closure.

General anesthesia with endotracheal intubation and antibiotic prophylaxis were routinely given to all patients before the procedure. Endotracheal intubation was removed in the operating or recovery room after the procedure. Patients were transferred to the general ward after their condition stabilized.

In group A, heparin (100 IU/kg) was given to all patients before femoral venous puncture. All procedures were performed using TTE/TEE monitoring and X-ray guidance in cardiac catheterization room. The size and morphology of defects were evaluated by TTE/TEE. The procedure was usually performed in accordance with the standard protocol, whereas several cases of pulmonary vein deployment technique could be used to facilitate the transcatheter closure of large secundum ASDs.

The intraoperative device closure technique used in group B has been described in previous studies [10‐15]. In brief, the TEE probe was inserted after general anesthesia with endotracheal intubation was given to patients, and ASD closure was performed by only using TEE guidance. A minimally invasive incision (2 cm to 3 cm) was made in the right anterior third or fourth intercostal space of the right sternal border. Then, a mini-retractor was used in this manipulation incision. The pericardium was opened and cradled to suspend the heart. A 5–0 prolene purse–string suture was stitched on the right atrium. A secure suture of 3–0 prolene was passed through the selected device under the microscrew. Then, the selected device was loaded into a delivery sheath. Heparin (100 IU/kg) was administered to all patients before the loaded sheath was inserted into the purse-string suture of the right atrium. The loaded sheath was passed through the ASD into the left atrium under TEE guidance. The left umbrella folder, waist, and the right umbrella folder were then deployed to close the ASD. After the device was successfully implanted, suitable views were used to assess the position and possible impacts of the occluder on surrounding structures. Color doppler flow was used to observe atrioventricular valve function, coronary sinus return, systemic/pulmonary venous return, and residual shunting. After the occluders were implanted in place, they were released while keeping the secure suture connected with the device for 3–5 min to observe the device position. The secure suture was then removed with the purse-string tied. The minimally invasive incision was closed without a drainage tube. Before the incision was closed, the aspirator was used to completely remove the fluid in the right chest cavity and pericardium and empty the air by expanding the lungs.

The day following the procedure, TTE, chest X-rays, and electrocardiogram were performed to check for device position, any residual shunt and arrhythmia, or other peri-procedural complications. All patients were discharged with aspirin medication (3–5 mg/kg body weight, orally) for 6 months. At 1, 3, 6, and 12 months, all patients underwent another TTE and electrocardiogram. Further annual clinical follow-up was scheduled for all patients. The main concern was whether translocation and thrombosis formation of the occluder, new arrhythmia, any residual shunt, aortic erosion, and other complications occurred.

As summarized in Table 1, statistically significant differences were observed in certain demographic data or preoperative clinical characteristics between the two groups. The mean age of patients was 28.3 ± 6.6 and 16.4 ± 8.0 months for groups A and B (P < 0.05), respectively. In group B, the mean body weight was significantly lighter than that in group A (9.7 ± 1.8 kg versus 12.9 ± 1.5 kg, P < 0.05). The proportion of complex ASDs (multiples or rims deficiency) in group B was significantly higher than that in group A.

Table 2 lists the intraoperative and post-operative results. Successful closure was obtained in 86 (97.7%) and 96 (98.0%) infants in groups A and B, respectively, with two failed cases in each group (2.3% vs 2%). An immediate device embolization in the RV occurred in group A, and the patients were referred to emergency surgery for device retrieval and surgical ASD closure. During rescue, device embolization caused cardiac arrest and blood flow interruption, causing minor brain complication that gradually improved after treatment. Percutaneous device closure was aborted in another child due to the oblique position of the implant in a relatively large defect. Intraoperative closure was successfully performed 1 week later in this child. In group A, the initially unstable implanted device was replaced by a second larger device in three patients. In group B, intraoperative device closure was abandoned in two patients because the devices were difficult to fix for the deficient inferior rims of ASDs. We immediately extended the incision to replace the device closure with a right chest surgical closure. In group A, the diameter of ASDs, as measured by TTE/TEE, ranged from 5 mm to 16 mm (9.1 ± 2.6 mm), whereas the size of the implanted occluder ranged from 8 mm to 22 mm (12.6 ± 3.5 mm). In group B, the corresponding data were 6 mm to 20 mm diameter (9.6 ± 2.7 mm, P = 0.233) and 8 mm to 24 mm device size (13.0 ± 3.0 mm, P = 0.404). In all patients, the ratio of the applied device size to body weight was determined. The device/weight reached 0.97 ± 0.23 and 1.35 ± 0.27 in groups A and B (P < 0.05), respectively, and the proportion of using a considerably large occluder (device/weight ratio ≥ 1.5) was significantly lower in group A than that in group B. Trivial residual shunts were documented immediately after the procedure in 3 and 8 patients in groups A and B (3.4% vs 8.2%, P = 0.17), respectively. The durations of the procedure and the postoperative hospital stay in group A were significantly shorter than those in group B (P < 0.05). Mean fluoroscopy time in group A was 4.6 ± 7.9 min.

Among the 184 patients who had a successful ASD closure, follow-up data were available for 83 (96.5%) and 91 (94.8%) patients in groups A and B, respectively. Three (3.5%) and five (5.2%) patients in groups A and B were lost to follow-up, respectively. After a median follow-up duration of 46 months (mean: 52.1 ± 26.5 months; range: 14–102 months) in group A and 58 months in group B (mean: 58.1 ± 28.0 months; range: 13–107 months), no death, thromboembolic events, aortic erosions, occluder dislodged, and other major complications were observed. At present, symptoms either resolved completely or improved significantly for all symptomatic infants. One patient in group B had trivial shunts at the 3-month follow-up evaluation, and none exhibited any residual shunts at the 6-month follow-up evaluation. In mid-term to long-term observation, mild mitral valve insufficiency occurred in one patient in group B. Pulmonary hypertension resolved in both groups as estimated by echocardiography, and patients with failure to thrive gained weight appropriately for their age. In both groups, the structure and hemodynamic parameters significantly improved during follow-up. RV end-diastolic anteroposterior dimension and the ratio of right ventricular to left ventricular end-diastolic transverse dimension (RV/LV) were recorded before ASD closure. RV dilation was present in all patients. RV dimension and RV/LV decreased in all patients at the two-year follow-up evaluation (Fig. 1).

This study was a single-center research performed on selected cases. This work was based on a relatively small series of patients, and the follow-up period was 55.3 ± 27.3 months. A multicenter, prospective, randomized trial is convictive. In addition, 4.3% of all patients were lost to follow-up. Further studies are required to establish additional long-term results in a large patient population.