Head-up tilt test ¹⁴

Drugs which could influence autonomic nervous system activity were stopped for at least 5 half-lives, and children were kept fasting for 4 hours before the head-up tilt test, which was conducted in a quiet and dim environment. Electrocardiography, heart rate, and blood pressure were monitored continuously by tilt bed system (Standard Medical Technology, Beijing, China). First, children were positioned flat on the bed for 10 minutes, after which the bed was tilted to 60°for 45 minutes or until a positive response occurred.

The positive diagnostic criteria in head-up tilt test were the following: ① decrease of blood pressure; ② drop of heart rate; ③ sinus arrest; or ④ second-degree or greater atrioventricular block and asystole persisting for > 3 seconds. On the basis of the diagnosis of vasovagal syncope, if cardiac arrest lasted for ≥ 3 seconds in association with syncope, diagnosis of malignant vasovagal syncope could be made. Sinus node dysfunction should be excluded. Clinically, it is feasible to further evaluate the function of the sinus node by electrocardiogram exercise stress test.

Pre-ablation preparation

All children were asked to discontinue drugs for at least 5 half-lives. All the procedures were performed under local anesthesia. The changes of electrocardiogram, blood oxygen saturation and blood pressure were continuously monitored during the procedure. The filtering range of electrocardiogram is 30–500 Hz and the speed is 100 mm/s. Electrodes were place in the coronary sinus and the right ventricle from the subclavian vein and femoral vein access. Modified atrial septal puncture was performed under X-ray fluoroscopy. ^{Reference Yao, Ding and Chen15} The 3-dimensional geometry of the left atrium was conducted under the guidance of the En site-Navx Velocity 5.0 system (Abbott) or Carto3 Version 6.0 system (Johnson & Johnson). Target mapping and ablation were performed in the left atrium using a cold saline irrigated-tip ablation catheter. High frequency stimulation mapping was performed on the endocardial surface of the left atrium to search for positive response points and radio-frequency ablation. During the operation, ventricular pacing was automatically output by a temporary pacemaker to avoid ventricular electrode connection with an EP4 simulator or high frequency stimulation by ventricular electrode output accidentally caused by the operator. Power and temperature were set with upper limits of 40 W and 60°C, respectively. The end point of the ablation procedure was taken as lack of vagal response induced with repeat high-frequency stimulation.

Catheter ablation of ganglionated plexus in the left atrium guided by high-frequency stimulation

Each ganglionated plexus was ablated by trial discharge methods. First, radiofrequency energy was released in the anatomical location area of each ganglionated plexus for 10 seconds. If the vagal response was triggered, including sinus arrest, atrioventricular block, ventricular escape, and the average R-R interval was extended by 50 per cent, continuous discharge was repeated in this area for at least 30 seconds, until the vagus response disappeared. On the contrary, if there was no vagal reaction within 10 seconds, the discharge was stopped and the ablation of adjacent parts attempted again under the guidance of three-dimensional mapping system. The end point of autonomic ganglion plexus ablation under anatomic guidance was that no vagal reaction was caused by five consecutive attempts of ablation of any ganglionated plexus anatomical location area, for instance, when ablation catheter was placed on the anterior wall of the right pulmonary vein in vasovagal syncope cardio-inhibitory children.

Protocol of follow-up after catheter ablation

After the operation, the patients stopped all drugs and were re-examined in the outpatient or inpatient department at 1 month, 6 months, and 1 year, until 5 years after the operation to record the recurrence of post-operative syncope in detail. Six months after the therapy, follow-up was conducted by a professional investigator by telephone or outpatient visits. As indicated before, symptom score was calculated before and at the end of the follow-up depending on the occurrence and frequency of syncope: 0 point indicated no syncopal event; 1 point = 1 event per month; 2 points = 2–4 events per month; 3 points = 2–7 events per week; and 4 points = > one per day. When the symptom score decreased by at least 1 point, the treatment was considered to be effective. ^{Reference Zheng, Sun and Qiao19} Head-up tilt test and 24-hour electrocardiogram were performed during the follow-up.

Calculation of deceleration capacity

The calculation of deceleration capacity was derived from the results of 24-hour electrocardiographic monitoring, and the recordings were digitised at 128 Hz and automatically processed with specific software based on the phase-rectified signal averaging algorithm. ^{Reference Tao, Chen, Li, Tang, Du and Jin9,Reference Zheng, Sun and Qiao19} First, the heartbeat intervals longer than the preceding interval are called anchors. R-R interval prolongations 5% were excluded to avoid errors caused by artefacts. Segments of the same size around the anchors were selected and aligned at the anchors and signals X within the aligned segments were averaged. Deceleration capacity is quantified using the following equation: deceleration capacity = (1/4)[X0 + X1−(X −1)−(X −2)]. X0 and X1 are the averages of the anchor points and the following R-R intervals, while X −1 and X −2 are the averages of the 2 R-R intervals preceding the anchor points. Deceleration capacity measures were calculated for the entire 24 hours from 8:00 to 23:00 (daytime deceleration capacity) and from 23:00 to 8:00 (night-time deceleration capacity).

Previous studies proposed that deceleration capacity reflected the cardiac parasympathetic tension. ^{Reference Tu, Wu and Hu13}

Statistical analysis

Data analysis was performed using SPSS version 22.0 (IBM, Armonk, New York). Data were expressed as mean ± SD, and the Shapiro–Wilk test was applied to evaluate the normality of the distribution of continuous data before statistical analysis. Non-normal distribution variables are expressed in quartile by median (Q1, Q3). The paired t test was applied to compare the symptom scores and deceleration capacity values between pre-ablation and post-ablation.