Journal international de recherche cardiovasculaire

Cardiac MRI Volumetric Assessment by Short-Axis has Better Reproducibility than the Axial Orientation in Functionally Single Right Ventricle Hearts Prior to Bidirectional Cavopulmonary Anastomosis

Sanam Verma, Nee S Khoo, Michelle L Noga, Kimberley A Myers, David J Patton and Edythe B Tham

Cardiac MRI Volumetric Assessment by Short-Axis has Better Reproducibility than the Axial Orientation in Functionally Single Right Ventricle Hearts Prior to Bidirectional Cavopulmonary Anastomosis

Background: Cardiovascular Magnetic resonance imaging (CMR) is the standard for assessment of ventricular volumes and function. However, few reports exist regarding methodology for measurements of functionally single right ventricular (RV) volumes in congenital heart disease. This study aims to determine which imaging plane, the short axis oblique (SAO) or the axial orientation (AX), provides greater reproducibility in evaluation of RV volumes and function in patients with hypoplastic left heart syndrome (HLHS) prior to bidirectional cavopulmonary anastomosis (BCPA).

Methods: CMR was performed under general anesthesia in 23 patients (5 ± 3.4 months) with single RV prior to BCPA to obtain ventricular volumes from the axial and/or short-axis cine orientations. Postprocessing (cmr42 - Circle Imaging, Calgary, Alberta) was performed by two independent observers to obtain end-diastolic (EDV), end-systolic (ESV) volumes and ejection fraction (EF) in both SAO (n=23) and AX (n=16) planes. Absolute differences (mean ± SD), repeatability values, intraclass correlation coefficients (ICC), coefficient of variation, and Bland Altman plots were used to assess the reproducibility between methods, and interobserver and intraobserver variability.

Results: Comparisons of SAO versus AX volumes revealed small absolute differences, with better repeatability for ESV (4.6 ml) compared to EDV (8.5 ml), and high agreement between the 2 methods. Interobserver variability showed high correlations and narrow limits of agreements for SAO and AX measured ESV. However, AX EDV had poorer repeatability (11 ml) and wider limits of agreement than SAO EDV (6.4 ml). Intraobserver testing showed higher correlations for all measurements, but SAO EDV (2.7 ml) showed better repeatability than AX EDV (5.7 ml).

Conclusion: CMR assessment of single RV volumes showed better inter and intra-observer reproducibility for the SAO than the AX methods suggesting that the addition of an axial stack does not provide any benefit over the short axis orientation in the assessment of single RV volumes and function.