Human spinal cord transplantation : experimental and clinical application

Sammanfattning: The main objectives of this Thesis was to further characterize the in situ human first trimester spinal cord tissue and to evaluate the survival, integration and possible neurological effects of solid human embryonic spinal cord grafts transplanted to cavities in the spinal cord of rat and man. Human-specific neurofilament and Thy-1 positive fibers were observed already at 5 weeks of gestation in the in situ human spinal cord, the earliest stage at which donor tissue was used for intraspinal transplantation. Radiating glial fibrillary acidic protein (GFAP) immunoreactivity (IR) suggestive of radial glia was first observed at 7.5 weeks of gestation. Binding sites for [3H]AMPA, [3H]kainate and [3H]MK-801 as well as immunoreactive bands corresponding to the NMDA receptor subunits NR1, NR2A, NR2B, NR2C and NR2D were demonstrated by receptor autoradiography and immunoblotting, respectively, already at 4-7 weeks of gestation and increased continuously during the first trimester. This indicated ionotropic glutamate receptor expression in the human spinal cord already during embryogenesis. Human leukocyte antigen (HLA) class I expression was observed in 5-17% and class II in 0-9% of first trimester human spinal cord cells. After 8 days in culture with [gamma]-interferon, >87% of the spinal cord cells expressed HLA class II. However, mixed cultures of human adult peripheral lymphocytes and immature human spinal cord cells, showed no induction of lymphocyte proliferation prior to or after [gamma]-interferon exposure in culture. In conclusion, non-immunogenic expression of HLA antigens was observed in the human first trimester spinal cord. Solid grafts of human embryonic spinal cord showed >94% survival rate after implantation to chronic and acute unilateral spinal cord cavities in the immune-incompetent rodent. However, 6 months after transplantation the graft volume was significantly larger in the chronic compared to the acute cavity group. The noduli-shaped grafts grew to fill the volume of the unilateral cavity in a restrictive manner and continued its human-specific neural differentiation expressing humanspecific neurofilament and Thy-1 IR fibers. The graft-host fiber integration was similarly substantial in the grafted chronic and acute cavity group. At 6 months, the grafted groups showed significantly lower degree of GFAP IR at the graft-host border, compared to that of the unilateral cavity alone group. Furthermore, grafted groups presented less deformations of the central canal compared to that of the non- grafted group. The open-field locomotor function as evaluated with the BBB scale showed no significant difference between the grafted and non-grafted groups, suggesting neither negative nor beneficial effects by the human spinal cord grafts in this lesion model. Solid human spinal cord grafts were implanted, in conjunction with conventional neurosurgical intervention, in three patients with progressive post-traumatic cystic myelopathy. The aim was to fuse the cyst walls and substitute for tissue loss with the grafts in order to reduce the risk of reexpansion of the cysts. For the first time, MRI-verified survival of transplanted solid human spinal cord tissue was observed after transplantation to man. The solid human spinal cord grafts obliterated the cysts at the site of implantation, as detected with MRI. Halted progression of the pre-operative neurological symtoms and modest neurological improvement showed that the surgical intervention was successful.