Abstracts

Rationale: Rasmussen encephalitis (RE) is a neurological disease of suspected auto-immune origin. It is characterized clinically by refractory partial epilepsy and a progressive hemisyndrome. The presence of antibodies targeting GluR3 receptors in some patients and the response of others to different forms of immune therapy support an underlying auto-immune mechanism. However, the exact mechanism of this catastrophic epilepsy has not been elucidated. The pathological hallmarks described by Dr Robitaille include in the active stage infiltration of T lymphocytes, microglial nodules, gliosis with or without neuronophagia. However, the exact mechanisms by which the immune system could be responsible of the disease are still unknown. Humanized mice have been used to study the pathophysiology of auto-immune disorders such as acute nephrotic syndrome by injecting lymphocytes of affected children in immunocompromised mice that can not reject the human cells and will permit the immature stem cells to proliferate and differentiate into all the lymphoid and myeloid immune populations. We propose that this could be a powerful avenue to investigate the pathophysiolgy of RE. Methods: We engrafted adult NOD/SCID/IL2rgnull mice with CD34+ stem cells isolated from the peripheral blood of 3 RE patients and 3 normal controls. Animals were observed in the immunology laboratory for two weeks to confirm engraftment prior to being transferred to our video-EEG monitoring unit. They were monitored for 5-10 days and then sacrificed by perfusion with paraformaldehyde 4%. Results: Mice injected with RE patient cells demonstrated generalized convulsions in 2/3 cases and interictal discharges in all. No EEG abnormalities and convulsions were observed in controls. Moreover, the brain of mice with RE cells showed the presence of human lymphocytes and severe gliosis, as observed in the human disease while only mild or no gliosis was observed in controls. Conclusions: These results demonstrate that injection of cells of RE patients in NOD/SCID/IL2rgnull mice is sufficient to induce in part the clinical and pathological properties of RE. This new mice model could be useful to elucidate the mechanisms of the disease and can serve for the development of new therapeutic avenues.