Animal model of intramedullary spinal cord glioma using human glioblastoma multiforme neurospheres. Laboratory investigation

Object. Advances in the diagnosis and management of patients with spinal cord tumors have been limited because of the rarity of the disease and the limitations of current animal models for spinal cord glioma. The ideal spinal cord tumor model would possess a number of characteristics, including the use of human glioma cells that capture the growth pattern and local invasive nature of their human counterpart. In this study, the authors' goal was to develop a novel spinal cord tumor model using a human neurosphere cell line. Methods. Eighteen female athymic rats were randomized into 3 experimental groups. Animals in the first group (6 rats) received a 3-ml intramedullary injection containing DMEM and were used as controls. Animals in the second group (6 rats) received a 3-ml intramedullary injection containing 100,000 glioblastoma multiforme (GBM) neurosphere cells in 3 ml DMEM. Animals in the third group (6 rats) received a 3-ml intramedullary injection containing 9L gliosarcoma cells in 3 ml DMEM. Functional testing of hindlimb strength was assessed using the Basso-Beattie- Bresnahan (BBB) scale. Once the functional BBB score of an animal was less than or equal to 5 (slight movement of 2 joints and extensive movement of the third), euthanasia was performed. Results. Animals in the GBM neurosphere group had a mean survival of 33.3 ± 2.0 days, which was approximately twice as long as animals in the 9L gliosarcoma group (16.3 ± 2.3 days). There was a significant difference between survival of the GBM neurosphere and 9L gliosarcoma groups (p < 0.001). None of the control animals died (p < 0.001 for GBM neurosphere group vs controls and 9L vs controls). Histopathological examination of the rats injected with 9L gliosarcoma revealed that all animals developed highly cellular, well-circumscribed lesions causing compression of the surrounding tissue, with minimal invasion of the surrounding gray and white matter. Histopathological examination of animals injected with GBM neurospheres revealed that all animals developed infiltrative lesions with a high degree of white and gray matter invasion along with areas of necrosis. Conclusions. The authors have established a novel animal model of spinal cord glioma using neurospheres derived from human GBM. When injected into the spinal cords of athymic nude rats, neurospheres gave rise to infiltrative, actively proliferating tumors that were histologically identical to spinal cord glioma in humans. On the basis of their results, the authors conclude that this is a reproducible animal model of high-grade spinal cord glioma based on a human GBM neurosphere line. This model represents an improvement over other models using nonhuman glioma cell lines. Novel therapeutic strategies can be readily evaluated using this model.