Article Abstract

Multimodality cardiac computed tomography angiography and magnetic resonance with clinical-grade scanners provide robust assessment of cardiac morphology and function in rabbits

Authors: Gitsios Gitsioudis, Maximilian Nunninger, Anna Missiou, Peter Wolf, Hugo A. Katus, Grigorios Korosoglou


Background: Non-invasive computer tomography (CT)- and magnetic resonance (MR)-based cardiac imaging still remains challenging in rodents. To investigate the robustness of non-invasive multimodality cardiac imaging in rabbits using clinical-grade CT and MR scanners.
Methods: A total of 16 rabbits (2.7–4.0 kg) serially underwent cardiac-gated imaging using a clinical-grade 256-row CT and a 1.5 Tesla MR-scanner at baseline and at 4-month follow-up (16±1 weeks). Image analysis included image quality (5-grade scale), left ventricular (LV) volumes, LV stroke volume, LV diameters, LV wall thickness and ejection fraction (LVEF).
Results: Cardiac MR (CMR) and CT angiography (CTA) provide images with an overall good image quality (excellent or good quality: CMR 82% vs. CTA 78%, P=0.68). Linear regression analysis demonstrated a good correlation of all diameters (diam.) and volumes (vol.) as assessed by CTA and CMR (diam.: r=0.9, 95% CI: 0.8–0.9; vol.: r=0.8, 95% CI: 0.6–0.9; P<0.0001 for both). CTA-based volumetric analysis revealed slightly higher LVEF values as compared to CMR (CTA: 64%±1%, CMR: 59%±1%, P=0.002). Analysis of inter-/intra-observer agreement demonstrated excellent agreements for diameters (CMR: 98.5%/98.7%; CTA: 98.2%/97.4%) and volumes (CMR: 99.9%/98.8%; CTA 98.7%/98.7%). Finally, serial CMR- and CTA-based assessment of cardiac diameters and volumes delivered excellent intersession agreements of baseline versus follow-up data (diam.: CMR: r=0.89; CTA: r=0.92; vol.: CMR: r=0.87; CTA: r=0.96, P<0.0001 for all).
Conclusions: Multimodality non-invasive assessment of cardiac function and aortic hemodynamics is feasible and robust in rabbits using clinical-grade and MR and CT scanners. These imaging modalities could improve serial cardiac assessment for disease monitoring in preclinical settings.