Cardiovascular magnetic resonance imaging systematically overestimated epicardial adipose tissue volume compared to cardiac computed tomography (median 89.8ml vs 73.2ml, p<0.001; mean bias +13ml).
Observational (n=137)
Does CMR provide comparable and reproducible epicardial adipose tissue volume quantification compared to CCT in patients with severe aortic stenosis?
While both CMR and CCT provide highly reproducible measurements of epicardial adipose tissue volume, CMR systematically overestimates volumes compared to CCT, indicating the modalities should not be used interchangeably for longitudinal monitoring.
Mean Difference: 13 (95% CI 38–64)
Absolute Event Rate: 89.8% vs 73.2%
p-value: p=<0.001
Abstract Background Increased epicardial adipose tissue volumes (EATV) exert adverse effects in cardiovascular disease. Emerging pharmacological strategies targeting elevated EATV may potentially mitigate these effects. Accurate quantification is therefore essential for reliable diagnosis and therapy monitoring. Cardiac computed tomography (CCT) is considered the reference standard for EATV quantification due to its high spatial and isotopic resolution. Cardiovascular magnetic resonance imaging (CMR) presents a radiation-free alternative. However, existing data on EATV derived from CCT and CMR lacks comparability as intermodality agreement and reproducibility have not been systematically investigated. Purpose This study aims to evaluate the reproducibility and intermodality agreement of EATV quantification using CMR compared to the reference standard CCT. Methods A total of 141 patients with severe aortic stenosis underwent paired CCT and CMR as part of a prospective clinical study. EATV quantification on CCT was performed by manual definition of the visceral pericardium and quantification of voxels within a threshold of -190 to -30HU. CMR-based quantification was carried out by manual volumetry and delineation of the visceral pericardium and epicardium on end-diastolic short-axis cine stacks. Details on acquisition parameters are compared in Table 1. Intermodality agreement was assessed using Spearman correlation and Bland-Altman analysis. Reproducibility was assessed in 20 patients per modality using the intraclass correlation coefficient (ICC) and coefficient of variation (CoV). Results A total of 137 patients with high-quality and complete imaging datasets were included for final analysis (78±6 years, 37,2% female). Median EATV measured by CMR was higher compared to CCT (89.8ml (73.0-113.5) vs. 73.2ml (56.8-101.7), p0.001), showing a mean bias of +13ml (95% CI: 38-64ml). As displayed in Figure 2, EATV of both modalities showed a moderate positive correlation (ρ=0.672, p0.001). EAT volume measurements showed excellent reproducibility for both modalities. CCT had very high intra- (ICC 0.997, CoV 8.8%) and inter-observer agreement (ICC 0.991, CoV 5.8%). CMR also showed good reproducibility (intra-observer ICC 0.915, CoV 16.4%; inter-observer ICC 0.934, CoV 14.4%). Conclusion Assessments of EATV using CMR demonstrate similarly high reproducibility as those derived from CCT when performed within the same modality. However, while showing good correlation, CMR systematically overestimates EATV compared to CCT. This highlights that for diagnostic purposes and therapy monitoring, measurements should not be directly compared across modalities, and follow-up should ideally use the same technique as the initial assessment.Figure 1 Figure 2
Gronwald et al. (Thu,) conducted a observational in severe aortic stenosis (n=137). Cardiovascular magnetic resonance imaging (CMR) vs. Cardiac computed tomography (CCT) was evaluated on Epicardial adipose tissue volume (EATV) (mean bias +13ml, 95% CI 38-64, p=<0.001). Cardiovascular magnetic resonance imaging systematically overestimated epicardial adipose tissue volume compared to cardiac computed tomography (median 89.8ml vs 73.2ml, p<0.001; mean bias +13ml).