Cine and tagged cardiovascular magnetic resonance imaging in normal rat at 1.5 T: a rest and stress study

Five rats were investigated for invasive left ventricle pressure measurements and five additional rats were imaged on a clinical 1.5 T MR system using a cine sequence (11–20 phases per cycle, 0.28/0.28/2 mm) and a C-SPAMM tag sequence (18–25 phases per cycle, 0.63/1.79/3 mm, tag spacing 1.25 mm). For each slice, wall thickening (WT) and circumferential strains (CS) were calculated at rest and at stress (2.5, 5 and 10 μg/min/kg of dobutamine).Good cine and tagged images were obtained in all the rats even at higher heart rate (300–440 bpm). Ejection fraction and left ventricular (LV) end-systolic volume showed significant changes after each dobutamine perfusion dose (p < 0.001). Tagged CMR had the capacity to resolve the CS transmural gradient and showed a significant increase of both WT and CS at stress compared to rest. Intra and interobserver study showed less variability for the tagged technique. In rats in which a LV catheter was placed, dobutamine produced a significant increase of heart rate, LV dP/dtmax and LV pressure significantly already at the lowest infusion dose.Robust cardiac cine and tagging CMR measurements can be obtained in the rat under incremental dobutamine stress using a clinical 1.5 T MR scanner.The clinical hallmarks of hibernating myocardium in chronic ischemic patients include regional contraction deficit while retaining an inotropic reserve during a low dose dobutamine challenge. Experimental models on rodents aiming to reproduce myocardial hibernation require an accurate quantification of the contractile reserve measured after low dose of dobutamine. Currently, two modalities echocardiography [1] and cardiovascular magnetic resonance (CMR) can provide cardiac function measurement in rodents. While echocardiography has been recently used to study cardiac response under dobutamine in normal rat [1], it is not routinely performed in rodent experiments. Echocardiography remains operator dependant and its reproducibility still needs to be large