The ventricular septal defect in tetralogy of Fallot: A multi-centre morphological and statistical shape analysis study

Abstract Background Ventricular septal defects (VSDs) in Tetralogy of Fallot (TOF) are often thought to be ‘large and unrestricted’ but experience suggests they have a more heterogenous morphology. Understanding the morphological variations and dimensions is the first step in modifying surgical closure or optimising potential device design. Purpose The objective of this study was to perform a comprehensive multi-center morphological analysis of the characteristics of the VSD in TOF. Additionally, we sought to perform a preliminary 2D shape analysis and principal component analysis (PCA) on standardised photos of the VSDs. Through this multi-dimensional approach, we aim to elucidate the morphological variations and structural characteristics of the VSD in TOF as a basis for future 3D shape analysis. Methods Of 495 specimens analysed, 460 were included in the morphological analysis. Standardised protocol was applied to assess the characteristics of the VSD: basic measurements of the height, width, inferior limb, and length of the inferior fibrous rim (if perimembranous) were taken and indexed to ventricular length. The degree of aortic override and presence of aorto-mitral continuity were also documented with double outlet right ventricle (DORV) being defined as more than 50% of aorta supported by the right ventricle. Statistical analysis was performed to assess variations in TOF morphology across different age groups and disease phenotypes. Standardised photographs of 130 VSDs were taken from the right ventricle and traced for PCA, in order to assess shape variations among the VSDs. Results Both the length of the inferior limb of the PM VSD and the length of the fibrous rim increased with ventricular length but the rate of change was lower for the fibrous rim. A chi-square analysis comparing the presence of double outlet right ventricle and the presence of a fibrous continuity between the aortic and mitral valves showed no association between these features (p = 0.89). The results of PCA in the indexed datasets showed that the largest variation was in the size of the VSD and then from the shape of the VSD where two distinct morphologies emerged: a classic V-shaped VSD and wide but shallow variant. Discussion Our study challenges conventional definitions and assumptions regarding the morphological characteristics of the VSD in TOF and suggests that components may change size at differing rates. Our study conclusively shows that DORV and TOF can co-exist and the definition of DORV should be made independently of the presence of aorto-mitral continuity. Contrary to the traditional view of TOF featuring large unrestricted VSDs, our findings suggest wide shallow VSDs are common. These findings challenge paradigms and underscore a necessity for the continuing morphological re-evaluation of TOF utilising large datasets with a diverse phenotypic spread.