ObjectivesTo prospectively compare in healthy rats the effect of multiple injections of macrocyclic (gadoterate meglumine) and linear (gadodiamide) gadolinium-based contrast agents (GBCAs) on T1-weighted signal intensity in the deep cerebellar nuclei (DCN), including the dentate nucleus.Materials and MethodsHealthy rats (n = 7/group) received 20 intravenous injections of 0.6 mmol of gadolinium (Gd) per kilogram (4 injections per week during 5 weeks) of gadodiamide, gadoterate meglumine, or hyperosmolar saline (control group). Brain T1-weighted magnetic resonance imaging was performed before and once a week during the 5 weeks of injections and during 5 additional weeks (treatment-free period). Gadolinium concentrations were measured with inductively coupled plasma mass spectrometry in plasma and brain. Blinded qualitative and quantitative evaluations of the T1 signal intensity in DCN were performed, as well as a statistical analysis on quantitative data.ResultsA significant and persistent T1 signal hyperintensity in DCN was observed only in gadodiamide-treated rats. The DCN-to-cerebellar cortex signal ratio was significantly increased from the 12th injection of gadodiamide (1.070 ± 0.024) compared to the gadoterate meglumine group (1.000 ± 0.033; P < 0.001) and control group (1.019 ± 0.022; P < 0.001) and did not significantly decrease during the treatment-free period. Total Gd concentrations in the gadodiamide group were significantly higher in the cerebellum (3.66 ± 0.91 nmol/g) compared with the gadoterate meglumine (0.26 ± 0.12 nmol/g; P < 0.05) and control (0.06 ± 0.10 nmol/g; P < 0.05) groups.ConclusionsRepeated administrations of the linear GBCA gadodiamide to healthy rats are associated with progressive and persistent T1 signal hyperintensity in the DCN, with Gd deposition in the cerebellum in contrast with the macrocyclic GBCA gadoterate meglumine for which no effect was observed.