The purpose of this study was to develop a metastatic brain tumor model of breast cancer in the rat to monitor the natural history of the disease using a multimodal imaging approach with magnetic resonance imaging (MRI) scanner and in vivo bioluminescence imaging (BLI). MDAMB231BR cell line is a brain seeking metastatic breast cancer line that was stably transfected to express the firefly luciferase cDNA (LUC) and was magnetically labeled with ferumoxides protamine sulfate (FEPro) complex ex vivo for early monitoring by MRI. Nude rats underwent intracardiac infusion FEPro ferumoxides labeled MDAMB231BRLUC (231BRL) cells. FEPro labeled 231BRL cells injected rats developed multiple brain and spinal cord metastasis and all rats with brain lesion had multiple skeletal metastasis. Tumor cell infiltrations were also detected to the lung, lymph nodes, and spleen by cytokeratin immunohistochemical staining. Prussian blue positive breast cancer cells could be detected in animals up to 1 week following intracardiac injection of FEPro labeled cells. BLI demonstrated increase in luciferase photon flux activity in the brain and bones of the rats while MRI revealed numerous hypointense regions corresponding to FEPro labeled cells in the brain within the first 3 days following injection of cell. By week 1 the luciferase photon flux activity similar to background photon flux and MRI rarely detected FEPro labeled cells past week 1. By 2 weeks, all animals had brain and skeletal metastasis on BLI and the luciferase activity increased in intensity with time. MRI detected metastatic lesions in the brain as early as week 2 postinfusion of 231BRL cells. The development of this metastatic breast cancer model in the rat allows for the use of imaging techniques to monitor the temporal and spatial migration of tumor cells and evaluate novel treatment strategies. Quantifying the number of labeled stem cells in target tissues is of great importance to optimize dose and timing of cellular therapy. T2 weighted spin echo imaging has been used to quantify the amount of iron within tissue. We has employed a multi-gradient echo pulse sequence to calculate T2* relaxation times to improve the detection of labeled cells within the voxel. To determine if FE-Pro labeled cells could be detected in tissues without the hypointense regions, the metastatic breast cancer model described above was used and fitted the signal intensities from multiecho T2* sequence as a monoexponential decay to calculate T2* maps. After normalizing the T2* values for the total number of voxels and quartile analysis was perform on the histogram distribution of T2* values over the brain, it was possible to compare serial MRI studies over time. There was significant differences in quartile 1 in the number of brain voxels that contained labeled versus baseline on days 1-3 following injection. A significant difference also existed at weeks 1-3 weeks in quartiles 2 and 3 after injection of FE-Pro labeled cells compared to animals that received unlabeled tumor cells. These results suggest that it may be possible to determine the presence of FEPro labeled cells in tissues for longer periods of time using quantitative T2* imaging versus qualitative T2* weighted images.
