Effect of at Least 10 Serial Gadobutrol Administrations on Brain Signal Intensity Ratios on T1-Weighted MRI in Children: A Matched Case-Control Study

BACKGROUND. An association is recognized between linear gadolinium-based contrast agents (GBCAs) and intracranial gadolinium retention in children. The relation between macrocyclic GBCAs and gadolinium retention remains incompletely understood. OBJECTIVE. The purpose of this study was to assess whether 10 or more administrations of the macrocyclic GBCA gadobutrol are associated with increased signal intensity (SI) in the dentate nucleus (DN) and globus pallidus (GP) on unenhanced T1-weighted MRI of children and to explore clinical variables potentially associated with T1 hyperintensity. METHODS. The case group consisted of 25 children (13 boys, 12 girls; mean age, 7 ± 4 years; range, 2-18 years) who underwent at least 10 (mean, 15 ± 6; range, 10-34) contrast-enhanced MRI examinations exclusively with gadobutrol. The control group consisted of 25 age- and sex-matched patients undergoing MRI who had never been exposed to gadolinium. Two observers in consensus using a 3-point scale assessed visual T1 hyperintensity in the DN and GP. One observer placed ROIs on T1-weighted images to mark the DN, GP, middle cerebellar peduncle (MCP), and pulvinar of the thalamus bilaterally to compute mean DN-to-MCP and GP-to-thalamus SI ratios. SI ratios were compared between the macrocyclic GBCA and control groups. In the macrocyclic GBCA group, Pearson correlation analysis was conducted between SI ratios and clinical variables. ROI measurements were repeated by the original reader and an independent reader, and interobserver and intraobserver agreement were computed by means of Lin concordance correlation coefficient (ρ_c). RESULTS. No patient had visual T1 hyperintensity in the DN or GP. No significant difference between the macrocyclic GBCA and control groups was observed for DN-to-MCP SI ratio (0.95 ± 0.05 vs 0.95 ± 0.03; p = .67) or GP-to-thalamus SI ratio (1.05 ± 0.06 vs 1.04 ± 0.06; p = .65). In the macrocyclic GBCA group, no significant correlation was observed between DN-to-MCP SI ratio or GP-to-thalamus SI ratio and age (r = 0.355, p = .08; r = 0.167, p = .42), number of contrast-enhanced MRI examinations (r = 0.247, p = .23; r = 0.203, p = .33), mean time between examinations (r = 0.193, p = .36; r = 0.047, p = .82), or cumulative macrocyclic GBCA dose (r = 0.434, p = .07; r = 0.270, p = .19). Interobserver and intraobserver agreement was substantial for DN-to-MCP SI and GP-to-TH SI ratios (ρ_c = 0.931-0.974). CONCLUSION. Ten or more serial gadobutrol administrations were not associated with T1 hyperintensity in the DN or GP of children. CLINICAL IMPACT. Selection of gadobutrol as an MRI contrast agent may reduce risk of gadolinium retention in children. The findings may help guide practices for GBCA administration to children.