Ventricular Septal Defect Area by Three-Dimensional Assessment of Shunt Severity in Children

Ventricular septal defect (VSD) size is a determinant of shunt severity in children. When compared to three-dimensional echocardiography (3DE) and surgical measures, Two-dimensional (2D) echocardiography (2DE) underestimates VSD size.

The authors therefore hypothesized that VSD area obtained from 3DE could better discriminate shunt severity than VSD diameters. Inclusion criteria included infants with isolated VSDs. Exclusion criteria included multiple VSDs, patients in the neonatal period or born prematurely at #37 weeks, and patients with any other cardiovascular or extracardiac anomalies including idiopathic pulmonary hypertension, syndromic, and genetic anomalies. Also excluded single VSDs having several orifices opening into the right ventricle side were excluded.

End-diastolic frames (first frame of mitral valve closure) were used to measure two orthogonal VSD diameters using 2DE from the long- and short-axis views. Other measurements were obtained: Aortic annulus diameter (PLAX view), pulmonary-to-systemic shunt ratio (QP/QS), left ventricular end-diastolic diameter and the z-score was calculated.

A ratio of QP/QS > 2 was considered significant volume overload, and a ratio of systolic pulmonary pressure/systolic arterial pressure > 50% was considered pulmonary arterial hypertension (PAH).

A 3D full-volume acquisition was obtained from the apical four chamber view and analyzed offline. Ventricular septal defect area was obtained by planimetry at the end-systolic frames (first frame after aortic valve closure) (sVSDA) and at the end-diastolic frames.

Sixty patients with median age of 9.5 months old were included in this study. There were 24 patients (40%) with small shunt, 21 patients (35%) with moderate shunt, and 15 patients (25%) with large shunt. Left ventricular end diastolic diameter Z score was higher as the shunt increased in size.

3D VSD areas have better discrimination power of the shunt severity compared with VSD 2D and 3D diameters. The best discriminant parameter of the severity of the shunt was the sVSDA/AVA ratio:

Moderate shunt: sVSDA/AVA > 0.2 (sensitivity 90.48%, specificity 95.8%)

Large shunt: sVSDA/AVA ratio > 0.33 (sensitivity 93.3%, specificity 95.2%)

CONCLUSION

The authors concluded that VSD area measured by 3DE is more discriminant of the shunt severity than VSD diameters measured by 2DE or 3DE. In addition, the systolic area of the defect was more discriminant than the diastolic area. Finally, the found that the sVSDA/AVA ratio was the best way to discriminate shunt severity with excellent sensitivity and specificity.