From January 2013 to November 2014, 164 consecutive patients (97 males, 59.1%) with VSD and AVP were treated with transcatheter closure at our institution. The clinical and imaging outcomes of these patients were retrospectively reviewed. The degree of AVP was evaluated by aortic root angiography and the degree of AR was evaluated by transthoracic echocardiography (TTE).

The morphology and function of the aortic valve were investigated in all subjects who met the following inclusion criteria: (i) diagnosis of PmVSD with AVP by cardiac catheterization; and (ii) age ≥ 24 months. The exclusion criteria were as follows: (i) PmVSD with AVP combined with other congenital heart diseases; (ii) VSD with a right to left shunt; (iii) bicuspid aortic valve; and (iv) other heart malformations requiring surgical treatment or residual shunt after VSD surgical repair.

This retrospective study was approved by the Ethics Committee of Guangdong Provincial People’s Hospital (No. GDREC2016427A). Individual written informed consent was obtained from each patients’ guardian before intervention.

The technique was performed as described previously [5,6]. The procedure was under the monitoring of TTE and fluoroscopy. After vascular access was established through the right femoral artery and vein, right and left cardiac catheterization and aortic root angiography were performed at the left anterior oblique 60° and cranial 20°. The location, shape, and size of the VSD, the degree and morphology of AVP and the relationship between the VSD and AVP were confirmed. A 5-F cut pigtail angiographic catheter was sent into the left ventricle and passed through the VSD. A 0.032-inch guide wire was fed along the 5-F catheter into the superior vena cava, inferior vena cava or pulmonary artery. A snare pulled the guide wire from the femoral vein, forming a femoral artery-VSD-femoral vein loop. An appropriate delivery sheath was introduced into the femoral vein along the guide wire, through the VSD into the ascending aorta, and then back to the left ventricle. An occluder size 1–3 mm larger than the size of VSD measured by left ventriculography was selected. An occluder was advanced along the delivery sheath, then the left disc and the waist were deployed in the left ventricle, the right disc was deployed in the right ventricle. Finally, left cardiac catheterization, aortic root angiography and TTE were performed again to evaluate the position, shape of occluder, the presence of residual shunts and AR before releasing the device.

Symmetrical VSD occluders (Lifetech Scientific Corporation, Shenzhen, China), eccentric VSD occluders (Lifetech Scientific Corporation, Shenzhen, China), and the Amplatzer Duct Occluder II (ADO-II; AGA Medical Corporation, Minnesota, USA) were used in our study.

Aortic root angiography was performed at the left anterior oblique 60° and cranial 20° in all patients before and after transcatheter closure of the VSD.

The degree of AVP was evaluated according to aortic root angiography at the end of ventricular diastole. AVP was divided into three grades [7,8]: mild (buckling of the aortic cusp down the left ventricular outflow tract without right coronary sinus prolapse into the VSD), moderate (buckling of the cusp with obvious herniation and right coronary sinus prolapse into the VSD) and severe (prolapse of the cusp and right coronary sinus prolapse into the right ventricle through the VSD).

The shape of the AVP was measured by aortic root angiography at the end of ventricular diastole. A straight line was drawn horizontally through the tip of the left coronary valve, parallel to the junction of the aortic sinus and ascending aorta. The depth and width of the AVP were measured. The data were corrected according to the diameter of the pigtail catheter used during radiography (Fig. 1).

Clinical data were obtained from databases, outpatient or telephone follow-up. All patients were required to undergo TTE, electrocardiogram (ECG) and clinical evaluation before intervention, and then 1 month, 3 months, 6 months, and 1 year after intervention and annually thereafter. The aborted procedure included death, more than moderate AR, new arrhythmias requiring pacemaker treatment, residual shunts or other complications requiring surgical or interventional treatment during follow-up.

The grading of AR was evaluated by TTE using the ratio of jet width and left ventricular outflow tract diameter during ventricular systole. AR was divided into four grades [8,9]: trivial (ratio < 10%), mild (ratio = 10%–24%), moderate (ratio = 25%–49%) and severe (ratio > 50%).

Aortic root angiography was performed at the left anterior oblique 60°and cranial 20°.The shape of the AVP was measured by aortic root angiography at the end of ventricular diastole. A straight line was drawn horizontally through the tip of the left coronary valve, parallel to the junction of the aortic sinus and ascending aorta. The data were corrected according to the diameter of the pigtail catheter used during radiography. A is the depth of AVP and b is the width of AVP.

The results are shown as n (%), mean ± standard deviation or median and IQRs. Statistical analysis was performed with SPSS 25.0 software (IBM Corporation, Chicago, USA). The Shapiro–Wilk test was used to verify normal distribution. The Student’s t-test was used to compare the groups of continuous variables with normal distribution, and the Mann–Whitney U-test was used to compare the groups of continuous variables without normal distribution or rank variable. Differences were considered significant when two-tailed P values were <0.05.

Patients’ characteristics are shown in Tab. 1. The mean age was 55.0 months (range 24–330 months), and mean weight was 16.94 ± 9.02 kg. All patients were divided into the mild AVP group (n = 63), moderate AVP group (n = 89) and severe AVP group (n = 12) according to the morphology of the aortic valve examined by cardiac catheterization before transcatheter closure.

Of the 164 patients, 146 (89.0%) were successfully treated with VSD occluders, including 59/63 (93.7%) with mild AVP, 80/89 (89.9%) with moderate AVP and 7/12 (58.3%) with severe AVP. A trivial to small residual shunt was found in 4 patients (6.8%) with mild AVP, in 4 patients (5.0%) with moderate AVP and in 1 patient (14.3%) with severe VSD. No deaths, hemolysis, or pericardial tamponade occurred during the perioperative period (Tab. 2). All patients with unsuccessful interventional therapy were admitted to pediatric cardiac surgery for surgical treatment (Fig. 2).

Of the 146 patients with VSD and AVP, the degree of AVP was ameliorated or even disappeared in 39 (26.7%) patients, and remained unchanged in 103 (70.5%) patients after the intervention. Two patients (3.4%) with mild AVP increased to moderate AVP, and one (1.2%) patient with moderate AVP increased to severe AVP after the intervention.

In patients with VSD and mild AVP (n = 59), the depth of the AVP was reduced from 1.27 ± 0.33 mm to 0.94 ± 0.53 mm (P < 0.0001), and the width of the AVP was reduced from 3.84 ± 1.32 mm to 2.82 ± 1.63 mm (P < 0.0001). In patients with VSD and moderate AVP (n = 87), the depth of the AVP was reduced from 2.24 ± 0.27 mm to 1.88 ± 0.56 mm (P < 0.0001), and the width of the AVP was decreased from 6.74 ± 1.66 mm to 5.56 ± 1.86 mm (P < 0.0001). In patients with VSD and severe AVP (n = 7), the depth of the AVP was reduced from 4.03 ± 0.68 mm to 2.92 ± 0.75 mm (P = 0.0015), and the width of the AVP was changed from 10.02 ± 1.74 mm to 9.20 ± 2.35 mm (P = 0.0503) (Fig. 3).

Of the patients with AR, 8 patients with mild AVP, 25 patients with moderate AVP and 2 patients with severe AVP successfully completed transcatheter closure of VSD. Two patients with VSD and AVP, accompanied by moderate AR were not successfully treated with transcatheter closure of VSD (Fig. 4).

Of the 35 patients who initially had trivial-to-moderate AR, the degree of AR was ameliorated or even disappeared in 25 (71.4%) patients, aggravated from trivial to mild AR in 1 (2.9%) patient, and remained unchanged in 9 (25.7%) patients after the intervention. In 111 patients without preprocedural AR, 1 (0.9%) patient had mild AR and 24 (21.6%) patients had trivial AR after transcatheter closure of VSD.

Of the 35 patients who initially had trivial-to-moderate AR, the degree of AR was ameliorated or even disappeared in 25 patients, aggravated from trivial to mild AR in 1 patient, and remained unchanged in 9 patients after the intervention. In 111 patients without preprocedural AR, 1 patient had mild AR and 24 patients had trivial AR after transcatheter closure of VSD.

Clinical follow-up ended in October 2020 and was complete in 137/146 (93.8%) of patients. The median (1st–3rd quartile) follow-up time was 70 months (54–77 months). During the follow-up period, no patients with AVP and AR required aortic valve intervention. Only one patient (0.7%) with moderate AVP increased from mild AR to moderate AR at one year after the intervention, and one patient (0.7%) with mild AVP increased from mild AR to moderate AR at two years after the intervention. The degree of AR in other patients remained stable at trivial-to-mild AVP or even disappeared during the follow-up period. No cases of valve perforation or aggravated regurgitation due to long-term attrition of the occluder occurred.