Echocardiography-guided percutaneous closure of perimembranous ventricular septal defects without arterial access and fluoroscopy

We enrolled 46 patients with isolated PmVSDs at Second Xiangya Hospital between January 1, 2014, and December 31, 2017 which included 26 male and 20 female patients (mean age, 6.5 years (range 4.2–12.0 years); mean body weight 22.1 ± 6.6 kg [range, 16.0–38.5 kg]; VSD diameter, 4.1 ± 0.6 mm [range, 3.2–5.0 mm]) (Table 1). The PmVSD were detected by pre-operation transthoracic echocardiography (TTE) (Vivid™ E9; GE Healthcare, Little Chalfont, United Kingdom) with color and Doppler interrogation from sub-xyphoid, apical, and parasternal views.

The patient selection criteria were as follows: 1) a left-to-right shunt whose Qp/Qs was ≥1.5 due to isolated PmVSD; 2) patient age between 2.0 years and 13.0 years; 3) patient weight > 12.5 kg; 4) maximum PmVSD diameter of ≤6 mm and minimum PmVSD diameter of ≥3 mm; and 5) subaortic rim (distance from the defect to the aortic valve) of > 2 mm. Patients with the following conditions were excluded: 1) more than a mild degree of aortic valve prolapse or regurgitation; 2) malalignment PmVSD; and 3) coexisting cardiac anomalies requiring correction during the same procedure.

Informed consent was obtained from the parents of all participants, and the study was approved by the Committee on Clinical Applications at the Second Xiangya Hospital. All patients were followed up with clinical examination, TTE, and electrocardiography at 1, 3, and 6 months, and 1 and 2 years after the procedure.

The VSD occluder (Fig. 1a and b; Shanghai Shape Memory Alloy Co. Ltd., Shanghai, China) was described in previous report [6]. In brief, this occluder is double-disk symmetrical concentric PmVSD occluder (S-pmVSO), and both discs were 2 mm larger than the waist. The waist was 4-14 mm long, with 3.5–4.5 mm diameter for its full length, corresponding to the size of the VSD in patients. The occluder size is selected according to the VSD diameter measured by TEE, with its diameter 2 mm larger than the VSD diameter.

The device is attached by a recessed microscrew to a delivery cable made of stainless steel (Fig. 1c). It is delivered through a 7F–8F long sheath. For introduction into the delivery sheath, the device is pulled into a special loader. The delivery system is also manufactured by Shanghai Shape Memory Alloy Co. Ltd., Shanghai, China.

The procedure was performed in a routine operating room under TEE guidance without fluoroscopy. All patients underwent general anesthesia with intubation in a supine position. TEE (Vivid E9 Dimension, bought from General Electric Company, Horten, Norway) was inserted into the middle esophagus. 9 T for patients less than 15 kg, 6 T for patients more than 15 kg. TEE studies were performed to evaluate the position, size, shape, direction of shunt flow, and the surrounding rims (particularly the subaortic rim) of the PmVSD prior to the procedure.

Femoral vein access was obtained using a 6-Fr sheath (Terumo Corporation, Tokyo, Japan). Heparinization was performed (125 U/kg) for anticoagulation. A 0.038­inch bent stiff guide wire (provided by Terumo Medical Corporation, Somerset, New Jersey, USA) was advanced from the inferior vena cava (IVC) via the right atrium (RA) into the superior vena cava (SVC) under TEE guidance using the bicaval view. A 5Fr JR 4 diagnostic catheter (provided by Cordis Corporation, Miami Lakes, Florida, USA) was subsequently advanced into RA over the wire. Then the guidewire was withdrawn progressively and the tip of catheter was tracked by the TEE short-axis view (Fig. 2A1 and A2). It may be necessary to adjust the direction of the catheter tip to advance through the tricuspid valve (TV) towards right ventricle (RV). The guidewire was then advanced into RV through the catheter (Fig. 2B1 and B2). The catheter was advanced into the RV over the guidewire, and the guidewire was then removed. Guided by TEE, along with the wire, the catheter’s direction was adjusted again to go through the outflow tract of RV towards main pulmonary artery (Fig. 2C1 and C2). The catheter was then gradually withdrawn and with the guidance of TEE, to make its tip point to the defect and the catheter advanced into the left ventricle (LV) (Fig. 2D1 and D2). The guidewire was guided slowly and gently through the VSD into the LV via the TEE four-chamber view (Fig. 2E1 and E2). We removed the vascular sheath and the right coronary catheter. The long delivery sheath (7F–8F) loaded with the dilator was advanced into the LV over the guidewire, and the dilator was then removed along with the guidewire, leaving just the tip of the sheath adjacent to the cardiac apex (Fig. 2F1 and F2). Thereafter, the occluder device was delivered along the sheath. Under TEE guidance, the sheath was then gently withdrawn to deploy the left disc in the LV (Fig. 2G1 and G2) and then pulled back until the disc was against the interventricular septum wall. The right side of the device was deployed by retracting the delivery sheath while applying slight tension on the cable (Fig. 2H1 and H2). Before the device was unscrewed, its correct and stable position was confirmed by TEE. Then TEE (Fig. 3a and b) and TTE (Fig. 3c and d) were performed to ensure that there were no residual shunts or aortic regurgitation. Bundle branch block was evaluated by electrocardiography during the procedure. If no postoperative complications (aortic regurgitation, complete atrioventricular block, or residual shunt) were found, both the loader sheath and the cable were withdrawn. The femoral vein was compressed for 5–8 min to stop the bleeding. A prophylactic antibiotic was administered during the procedure and on the day after. Moreover, aspirin (3 mg/kg/day) was routinely administered for 6 months.

If complications (aortic regurgitation, complete atrioventricular block, or residual shunt) were detected, the patients underwent open heart repair under CPB.

Data are expressed as percentages for nominal variables and mean ± standard deviation for continuous variables. Statistical Program for Social Sciences (SPSS) for Windows, version 24.0 (IBM, Armonk, NY, USA), was used for statistical analysis.

All procedures were carried out in a standard theatre without fluoroscopy backup, but with a CPB unit stand­by. The transcatheter occlusion procedure under TEE guidance was successful in 44 (95.7%) patients (Additional file 1: Video S1). PmVSD occlusion failed because of a residual shunt (width > 2 mm; flow rate > 3 m/s) in 2 patients (VSD diameter, 4.5 mm and 4.6 mm; subaortic rim, 3 mm) with no improvement even after using a 6-mm, 7-mm, and 8-mm occluder, and then they underwent open heart surgery.

Symmetrical VSD occluder with a size (waist length) of 6 mm or 7 mm was used in this study. The mean procedural duration, post-operative mechanical ventilation duration, intensive care unit (ICU) residence, and in-hospital durations were 28.2 ± 8.7 min (range, 12.0–42.0 min), 65.2 ± 5.6 min (range, 56.0–78.0 min), 2.1 ± 0.1 h (mean, 2.0–2.4 h), and 2.7 ± 0.2 d (range, 2.5–3.0 d), respectively. There was no need for blood transfusion in any of our patients, who underwent percutaneous device closure (Table 1).

The patients were followed up by clinical examination, echocardiography, and electrocardiography at 1, 3, and 6 months, and 1 and 2 years postoperatively. All patients survived without any peripheral vascular injury, acute complications, or severe adverse events (death, valve injury, complete atrioventricular block or embolism) during the early period or follow-up.

Complete atrioventricular blocks were not detected, whereas IRBBB was observed in 3 patients. A new IRBBB was discovered in 1 case during the first postoperative month, and the IRBBB resolved spontaneously and changed to normal rhythm in 3 patients during the third month of follow-up. All patients showed sinus rhythm in the period of 12–24 months follow-up.

Tricuspid regurgitation had decreased in 1 patient, whereas no new atrioventricular regurgitation was detected in the other patients at the first month of follow-up.

A trivial residual shunt (width, 1 mm; flow rate < 2.0 m/s) was detected in 1 patient on discharge. Fortunately, the residual shunt disappeared at the third month of follow-up (Table 1).

No other complications such as device dislocation, thrombosis, or obstruction of the left or right ventricular outflow tract were observed in the present study.