Atrial function after treatment of atrial septal defects in children under 5 years

Background Recently, the patient’s age of atrial septal defect (ASD) treatment has been gradually decreasing. However, the most appropriate age and treatment method remain controversial. We hypothesized that treatment of ASDs in patients under 5 years would be able to adequately normalize bi-atrial function over one year after treatment. The purpose of this study is to confirm the normalization of hypothesized atrial function. Methods Data of fifteen patients who underwent surgical ASD closure under 5 years of age (Operation group), 15 patients who underwent percutaneous ASD closure (Device group), and 15 age- and gender-matched normal control patients (Control group) were extracted from our echocardiographic data. Conventional 2D images and 2D speckle tracking method were used to evaluate bi-atrial function. Left atrial function, εS, εE, and εA showed no significant differences in the three groups. Indicators representing the right atrial function varied with the three groups. However, there were no significant differences in the global longitudinal strain of the right atrium between the Operation and Device groups.


Introduction
An atrial septal defect (ASD) is a relatively common disease of congenital heart defects. If the ratio of pulmonary blood flow and systemic blood flow through the ASD is more than 2:1 or if the patient has symptoms of ASD, treatment is performed by surgery or percutaneous trans-catheter closure. The purpose of treating ASD is to block the volume overload of the right atrium (RA) and right ventricle (RV) in order to reduce heart failure symptoms and prevent complications caused by secondary lung injury or arrhythmia caused by ASD. Infants and preschoolers rarely develop symptoms due to ASD, and more often they follow the natural course. However, hemodynamically significant ASDs require treatment during infancy and preschool ages. Recently, as percutaneous ASD closure has become more common, many cases of percutaneous device closure have been attempted except in early preschool ages because at that age, the patient's physique and vessel diameter is small, which increases the risk of various complications due to percutaneous intervention. Moreover, there is a risk of complications following surgical ASD closure. Despite these adverse factors, few studies have been conducted on how atrial function is restored in patients with ASD who underwent surgical or percutaneous device closure during the preschool age. The purpose of this study was to compare the atrial function of patients who underwent surgical ASD closure under the age of 5 years, those who underwent percutaneous device closure of ASD, and normal controls.

Population
Among patients who have undergone cardiac echocardiography using a GE Vivid-E95 or GE Vivid-E9 ultrasound scanner (GE Vingmed Ultrasound AS, Horten, Norway) within the last 3 years, fifteen patients who had undergone an ASD operation under 5 years of age at least 1 year ago (Operation group), fifteen patients who had undergone percutaneous device closure of ASD under 5 years of age at least 1 year ago (Device group), and fifteen healthy age-and gender-matched patients were selected (Control group). Clinical information of patients and controls included in the study was retrospectively obtained through a review of medical records. The institutional review board of Sejong general hospital approved this study and waived the need for informed consent due to the retrospective nature (Appraval number: 1964)

Echocardiographic data
There are three functions of the atria: the reservoir, conduit, and contraction functions. Left atrial (LA) volume and RA volume were measured by the area-length method using apical four-chamber viewobtained standard grayscale images. The maximum volume (V max ), minimum volume (V min ), and volume just before the P wave emerges from the surface electrocardiogram (ECG) (V preP ) of the RA and LA were measured by three pediatric experts (J.K., S.P., J.Y.). The calculated values are shown in Table 1. Table 1 Indicators of the three atrial functions Reservoir function Emptying volume = V max -V min Emptying fraction = 100-(V max -V min ) / V max Conduit function Passive emptying volume = V max -V preP Passive emptying fraction = 100 * (V max -V preP ) / V max Pump function Active emptying volume = V preP -V min Active emptying fraction = 100*( V preP -V min ) / V preP RA and LA 2D speckle tracking analysis was obtained using a grayscale apical four-chamber view.
Offline analysis of these images was performed using the EchoPAC software (GE Ultrasound AS, Horten, Norway), according to the current guidelines. Strain of the RA and LA was measured during peak systole (εS), early diastole (εE), and late diastole (εA; Fig. 1).

Baseline Characteristics
Each group (Operation, Device and Control) was age-and gender-matched (10 girls and 5 boys).
There were no significant differences in weight, height, and BSA between groups. At the time of echocardiography, the Operation group was on average 46.95 months after surgery and the Device group was 26.53 months after percutaneous ASD device closure. Most of the patients included in the Control group underwent echocardiography as part of an assessment for functional heart murmur.
None of the patients included in the study had definite arrhythmias or other systemic disease. The basic characteristics of the study populations are summarized in Table 2. There was no significant difference in age and the sex ratio between groups, and there was also no significant difference in weight, height, and BSA. Although there was a difference between the elapsed time after surgery in the Operation group and after percutaneous device closure in the Device group, this study analyzed echocardiography performed after more than one year in all patients.

Conventional echocardiographic indicators
There were no structural abnormalities nor pericardial or pleural effusion. There was no evidence of significant chamber enlargement or ventricular dysfunction. However, the Operation group had smaller systolic left ventricular dimension (LVIDs) than the Control group. There was no significant difference in mitral inflow Doppler pattern and mitral annular velocity measurements using tissue Doppler imaging (TDI) between groups. However, the tricuspid annular velocity measurement using TDI revealed significant differences between the Operation and Device groups (Table 3).  In our center, surgical ASD closure is more common than device closure if the child is under 3 years old or weighs less than 10 kg.
Unlike adult patients who have surgical or device closure of an ASD, in children, there are few cases of post-operative death and very few serious complications. Additionally, atrial fibrillation and pulmonary hypertension are rarely seen. Unlike adult patients, while the period of exposure to the left to right shunt is short, the hemodynamic significance of ASD required closure in children under 5 years also means that the amount of shunt is significantly larger to show clinical symptoms, which means that they are exposed to excessive volume overload. So far, there have been reports that the degree of RA and LA functional recovery varies according to the age at ASD treatment. 2−4 After the treatment of these ASDs, the question remains whether the function of the RA and LA can be normalized after some time. Indeed, the number of studies regarding atrial and ventricular function after ASD closure in childhood are very few. 5−12 In general, e' velocity is an indicator of the diastolic function of the ventricle and a' velocity is considered to be associated with the atrial function. As shown in Table 3, tricuspid e' and a' are significantly different between groups, and there are also significant differences between the Operation and Device groups, while mitral e' and a' show no significant differences. This is because ASD itself is a disease that mainly affects the right ventricle and RA. However, even after treatment for ASD, the LA and RA function of patients with ASD are reported to be inferior to those of normal subjects. 7, 13−14 To compare the reservoir, conduit, and pump functions, which are the three functions of the atria, V max , V min , V pre−P of the RA and LA were calculated using the obtained standard grayscale images of an apical four-chamber view by area-length methods. Although there are many limitations in this process, there were significant differences between groups in the LA reservoir function. On the other hand, the conduit and pump functions did not show significant statistical difference. However, the reason for the difference between groups may be an error in the process of comparing the calculated values with the assumption that the shape of the left atrium is constant. One unchanging principle is that both the Operation and Device groups have lower LA volumes than the Control group.
The Operation and Device groups showed significant differences in RA reservoir and conduit functions, but this difference was very small between the Device and Control groups. These results suggest that device closure of ASD can maintain conduit and reservoir functions of the RA relatively well, even at ages less than five years, compared to surgical closure. The reason for this may be related to scarring of the RA which inevitably occurs during surgical closure of ASD.
In the case of the RA active emptying fraction, which is the pump function of the RA, there was no significant difference between the Operation and Device groups, but the value was lower in each of those groups than in the Control group. These results may be attributed to the deterioration of interatrial wall motion occurring after device closure and scarring which remain at the RA incision site and interatrial septal patch.
2D speckle tracking analysis has been widely used to evaluate ventricular function, 15−16 but similar methods can be used to evaluate the atrial function. The εS, εE, and εA values obtained using EchoPAC software can be considered as indices representing the reservoir, conduit, and pump functions, respectively. The εS, εE, and εA of the LA did not differ significantly between groups.
Perhaps it is because the structural and functional changes of the heart due to ASD have relatively little effect on the LA.
In spite of the data obtained through the echocardiography conducted at least one year after the operation or device closure of the ASD, in the case of the RA, the εS, εE, and εA were all reduced in the Device group compared to the Control group. The Operation group showed significantly decreased εS and εE compared to the Control group. These results can be interpreted as a failure to normalize the LA and RA function even after one year of repair.
As a study conducted in adults reported that the RA strain during the reservoir phase is a predictor of atrial fibrillation development 17 , the possibility of future arrhythmias in the Operation and Device groups with decreased εS may be considered. Therefore, ASD patients treated under 5 years of age require long-term follow-up with the possibility of arrhythmia and reduced atrial function.

Study Limitations
This is a retrospective study with a low sample size. Since the area-length method was used to calculate the volume of the RA and LA, it may not accurately reflect the actual volume. When measuring global longitudinal strain using 2D speckle tracking, the values measured by the three pediatric cardiologists were averaged and did not solve the problem of inter-and intra-observer variation.

Conclusion
After ASD treatment, the recovery of RA function was less than that of the LA. In children under five years, RA function was not normalized after more than one year after treatment for ASD. After ASD treatment, further follow-up of the RA and LA function is necessary.