Journal of Cardiovascular Magnetic Resonance (Mar 2019)

Validation of aortic valve 4D flow analysis and myocardial deformation by cardiovascular magnetic resonance in patients after the arterial switch operation

  • W. H. S. van Wijk,
  • J. M. P. J. Breur,
  • J. J. M. Westenberg,
  • M. M. P. Driessen,
  • F. J. Meijboom,
  • B. Driesen,
  • E. C. de Baat,
  • P. A. F. M. Doevendans,
  • T. Leiner,
  • H. B. Grotenhuis

DOI
https://doi.org/10.1186/s12968-019-0527-6
Journal volume & issue
Vol. 21, no. 1
pp. 1 – 10

Abstract

Read online

Abstract Background Aortic regurgitation (AR) and subclinical left ventricular (LV) dysfunction expressed by myocardial deformation imaging are common in patients with transposition of the great arteries after the arterial switch operation (ASO). Echocardiographic evaluation is often hampered by reduced acoustic window settings. Cardiovascular magnetic resonance (CMR) imaging provides a robust alternative as it allows for comprehensive assessment of degree of AR and LV function. The purpose of this study is to validate CMR based 4-dimensional flow quantification (4D flow) for degree of AR and feature tracking strain measurements for LV deformation assessment in ASO patients. Methods A total of 81 ASO patients (median 20.6 years, IQR 13.5–28.4) underwent CMR for 4D and 2D flow analysis. CMR global longitudinal strain (GLS) feature tracking was compared to echocardiographic (echo) speckle tracking. Agreements between and within tests were expressed as intra-class correlation coefficients (ICC). Results Eleven ASO patients (13.6%) showed AR > 5% by 4D flow, with good correlation to 2D flow assessment (ICC = 0.85). 4D flow stroke volume of the aortic valve demonstrated good agreement to 2D stroke volume over the mitral valve (internal validation, ICC = 0.85) and multi-slice planimetric LV stroke volume (external validation, ICC = 0.95). 2D flow stroke volume showed slightly less, though still good agreement with 4D flow (ICC = 0.78) and planimetric LV stroke volume (ICC = 0.87). GLS by CMR was normal (− 18.8 ± 4.4%) and demonstrated good agreement with GLS and segmental analysis by echocardiographic speckle tracking (GLS = − 17.3 ± 3.1%, ICC of 0.80). Conclusions Aortic 4D flow and CMR feature tracking GLS analysis demonstrate good to excellent agreement with 2D flow assessment and echocardiographic speckle tracking, respectively, and can therefore reliably be used for an integrated and comprehensive CMR analysis of aortic valve competence and LV deformation analysis in ASO patients.

Keywords