Minimally invasive closure of transthoracic ventricular septal defect: postoperative complications and risk factors

Retrospectively analyzed the clinical data all patients underwent transthoracic ventricular septal defect closure performed in the Department of Cardiothoracic Surgery, Children’s Hospital of Chongqing Medical University from January 2018 to January 2020. The study was conducted in accordance with the Declaration of Helsinki and was approved by the Ethical Committee of Chongqing Medical University. Written informed consent was obtained from the patients’ parents before inclusion in this study. The research data were anonymized, and personal identifiers were completely removed. Patients were selected according to the following indications: All patients were diagnosed by transthoracic echocardiography before surgery, VSD size was 3-8 mm, clear left to right shunt and with or without mild to moderate pulmonary hypertension. Exclusion criteria: ①The size of the VSD is greater than 8 mm or less than 3 mm; ②The age is less than 5 months; ③Combined with other types of complicated congenital ④Patients who refuse to undergo this operation.

We included a total of 209 patients. We collected the patient’s information from the hospital’s electronic medical record system, ① basic information: gender, age, weight ②preoperative data: defect size, defect type, with or without pulmonary hypertension and with or without combined with aortic valve prolapse ③intraoperative data: the size of the defect, the size of the occluder, the type of the occluder, other operations during the operation, the surgeons and the operation time ④postoperative data: residual shunt, postoperative hospital stay, postoperative serious adverse events ⑤follow-up data.

The types of VSD defects can be divided into three groups: perimembrous defect, subarterial defect and muscular defect. The sonographer will initially assess whether the child has pulmonary hypertension before surgery. Pulmonary artery systolic pressure (PASP) > 25 mmHg is considered to have pulmonary hypertension. Postoperative outcomes and complications were evaluated according to international standards [5], including postoperative hospital stay, residual shunt and serious adverse events. For severe residual shunt, reoperation, reintubation, wound infection, post-pericardiotomy syndrome, complete atrioventricular block, unplanned second surgery, heart block requiring permanent pacemaker, circulation failure and death are considered serious adverse events.

Transthoracic cardiac color Doppler ultrasound and electrocardiogram were routinely reviewed 3–7 days after surgery, and cardiac color Doppler ultrasound and electrocardiogram were followed up in the outpatient clinic at 1, 3, 6, and 12 months after surgery. All children received oral aspirin 3–5 mg/kg daily for 3–6 months.

After anesthetized the child in the operating room, took the supine position, and evaluated the position, size, upper and lower edges of the defect and the aortic valve by TTE. A small incision of 2 to 4 cm in the middle of the skin (the skin incision of the perimembrous VSD was located in front of the lower end of the sternum and entered the mediastinum through the lower part of the sternum. The skin incision of the subarterial VSD was located in front of the middle sternum, and the skin incision was pulled to the left to enter the mediastinum through the third intercostal space on the left edge of the sternum.) Cut and suspend the pericardium. After 0.5 mg/kg heparinization, TEE guided the positioning of the right ventricular surface purse-string suture, selected domestic eccentric or equilateral occluder (Shanghai Shape Memory Alloy Co, Ltd), trocar punctured the right ventricular wall, removed the needle core, and under TEE monitoring, guided the wire from the right ventricle through the ventricular septal defect to the left ventricular cavity, placed the sheath, released the left ventricular surface umbrella disk, gently pulled back to make the left umbrella disk close to the ventricular septum, retracted the sheath to the right ventricular cavity, released the occluder waist and right ventricular surface umbrella tray. The “Mark” of the eccentric umbrella disc is far away from the aortic valve annulus. Subatrial ventricular septal defect or ventricular septal defect edge less than 2 mm from the aortic valve annulus choosed eccentric occluder, and the rest choosed equilateral occluder. For tunnel-type ventricular septal defects with inconsistent diameters of the left and right ventricular surfaces, the occluder is selected based on the right ventricular surface. TEE immediately evaluated the therapeutic effect, including residual shunt, function of the surrounding structure of the occluder, hemodynamics, stability of the occluder, pericardial effusion, heart rhythm, etc. If there is residual shunt, increase the occluder model. If it meets the quality control standards: no residual shunt, mild to moderate valve regurgitation, stable occluder, no complete atrioventricular block, no pericardial effusion, removed the occluder from the delivery device, and half protamine neutralized heparin. The surface of the right ventricle is ligated with a purse-string suture, stop bleeding and closed the chest. If it does not meet the quality control standards, transfer to surgery under cardiopulmonary bypass.

The Kolmogorov-Smirnov test is used to verify whether our data were normally distributed or not. For normally distributed data, we used the means and standard deviations to describe variables. For not normally distributed data, we used medians and ranges to describe variables. Frequencies and percentages were used for categorical data. The related risk factors of surgical complications were analyzed by univariate logistics regression. In order to avoid meaningful factors from being screened out, we set P < 0.1 as statistically significant. The statistically significant factors of univariate analysis were included in the multivariate logistics regression analysis, and the forward LR method is used. In the end, the statistically significant factors were independent risk factors. P < 0.05 means the difference is statistically significant. The statistical analyses were performed using SPSS version 24 (IBM, Armonk, NY, U.S.A).

A total of 209 patients received minimally invasive closure of transthoracic ventricular septal defect (Table 1). 55% were men, the median age of surgery was 29 months, the median weight was 12 kg, preoperative transthoracic ultrasound revealed 76.6% of the defects were perimembrous defects, and only 2.4% of muscular defects. Before surgery, 24.4% of patients had mild to moderate pulmonary hypertension. 39.2% of patients suffered from aortic valve prolapse, and 80.5% of them had mild aortic regurgitation.

A total of 7 surgeons were included in the study. 11% of patients underwent transthoracic atrial septal defect closure during the operation, 3% underwent arterial duct ligation during the operation, and 1 patient with diaphragmatic bulge underwent thoracoscopic surgery at the same time. The median size of the VSD defect was 5 mm, and the median size of the occluder used during the operation was 6 mm. Three patients used 2 occluders during the operation due to the presence of two defects, and 30.1% of the patients used eccentric occluders (Table 2).

After surgery, one patient was intubated for 95 h due to severe lung infection and supraventricular tachycardia, and another patient was intubated for 406 h due to severe pulmonary hypertension and severe lung infection. The average postoperative hospital stay was 7.88 ± 2.93 days, which was related to our hospital’s clinical path and postoperative health management. For those children whose hospital stay was longer than 9 days, these patients had complications after surgery, which we defined as postoperative Long hospital stay. A total of 27 patients were hospitalized for more than 9 days after surgery. The main reason for the prolonged hospital stay was infection (respiratory tract). A total of 33 patients had residual shunt after operation, and the residual shunt width was 1 to 2.3 mm, flow velocity 1.83 to 3.46 m/s. After 1–12 months of follow-up, except for 4 patients who were lost to follow-up, 2 patients persisted with residual shunt, but all had a small residual shunt < 2 mm, and the residual shunt of the remaining patients closed spontaneously. One patient had mechanical hemolysis after surgery and immediately underwent a second operation to remove the occluder. Two patients were re-operated 1 month and 10 months after surgery because of persistent moderate to severe aortic regurgitation. A large amount of pericardial effusion was found in 3 patients after the operation and received pericardiocentesis, and 2 patients developed a new atrioventricular block after the operation. And no patients died. Serious adverse events occurred in 3.3% of patients after surgery and follow-up.

Univariate analysis regression analysis was performed to determine the risk factors of long postoperative hospitalization, residual shunt and serious adverse events after surgery, and the meaningful factors of univariate analysis regression analysis were incorporated into the multivariate analysis regression analysis. Long postoperative hospitalization after surgery is defined as the length of hospital stay greater than 9 days. Due to the rarity of serious adverse events, we only conduct univariate analysis regression analysis on it. The analysis results of long postoperative hospitalization, residual shunt and serious adverse events are shown in Table 3. The size of VSD defect (OR: 1.494, 95% Cl: 1.108–2.013, P value: 0.008) is related to long postoperative hospitalization after surgery. The postoperative residual shunt is related to the size of the occluder (OR: 1.452, 95% Cl: 1.164–1.810, P value: 0.001). In univariate analysis, no risk factors related to serious adverse events were found.