3D Echo systematically underestimates right ventricular volumes compared to cardiovascular magnetic resonance in adult congenital heart disease patients with moderate or severe RV dilatation

Abstract

Background: Three dimensional echo is a relatively new technique which may offer a rapid alternative for the examination of the right heart. However its role in patients with non-standard ventricular size or anatomy is unclear. This study compared volumetric measurements of the right ventricle in 25 patients with adult congenital heart disease using both cardiovascular magnetic resonance (CMR) and three dimensional echocardiography.

Methods: Patients were grouped by diagnosis into those expected to have normal or near-normal RV size (patients with repaired coarctation of the aorta) and patients expected to have moderate or worse RV enlargement (patients with repaired tetralogy of Fallot or transposition of the great arteries). Right ventricular end diastolic volume, end systolic volume and ejection fraction were compared using both methods with CMR regarded as the reference standard

Results: Bland-Altman analysis of the 25 patients demonstrated that for both RV EDV and RV ESV, there was a significant and systematic under-estimation of volume by 3D echo compared to CMR. This bias led to a mean underestimation of RV EDV by -34% (95%CI: -91% to + 23%). The degree of underestimation was more marked for RV ESV with a bias of -42% (95%CI: -117% to + 32%). There was also a tendency to overestimate RV EF by 3D echo with a bias of approximately 13% (95% CI -52% to +27%).

Conclusions: Statistically significant and clinically meaningful differences in volumetric measurements were observed between the two techniques. Three dimensional echocardiography does not appear ready for routine clinical use in RV assessment in congenital heart disease patients with more than mild RV dilatation at the current time.

Figure 1

Off line processing of 3D echo data. Reconstructions are performed to generate images in 4 chamber, short axis and right ventricular inflow-outflow views. Contours are applied in all 3 planes at end systole and diastole and propagated in a semi-automatic fashion across all acquired time points. This allows generation of a time-resolved volume rendered image of the right ventricle (see Additional File 1: supplementary video file) and also a time volume curve from which functional data are obtained.

Figure 2

a-c - Comparative RV volumes & function by modality. Comparison of RV volumes and function between the mixed lesion (ToF and TGA) group and the CoA group. Boxes represent median and inter-quartile ranges and whiskers are the 95%CI.

Figure 3

a-i - Systematic differences between modalities for all patients combined. Groups comparison, correlation and Bland Altman analysis of difference in RV volumes and function as measured by both CMR and 3D echocardiography. Boxes represent median and inter-quartile ranges and whiskers are the 95%CI. Bland Altman plots demonstrate mean bias (dot-dash line) and 95% CI (dotted lines).