| Implanted biomedical devices play an important role in the treatment of central nervous system diseases and disorders, but are also subject to a foreign body response which has the potential to affect device function. For example, the recordings obtained from the cerebral cortex using penetrating microelectrode arrays have been shown to be sufficient in limited clinical trials to control computers or external devices in the field of neuroprosthetics, but such recordings are largely inconsistent in chronic applications. Following implantation, a local area of inflammation develops surrounding the implant, which consists of activated microglia/macrophages, astrocyte hypertrophy, and neuronal loss, and which appears to last for the lifetime of the implant. The focus of this dissertation was to investigate the local tissue response in rat cerebral cortex adjacent to microwires, silicon microelectrodes with different surface chemistries, and microelectrode designs with decreased surface area available for inflammatory cell attachment. In addition, the effect of the implant-associated foreign body response to hippocampal neurogenesis was elucidated. We observed that the chronic tissue response to implanted microwires and silicon microelectrodes is similar to other devices implanted in the brain, such as the deep brain stimulator and cerebrospinal shunt, and consisted of chronic inflammation and neuronal loss within the first 50 mum from the electrode-tissue interface. Local neuronal loss did not progress from 2 to 12 weeks, and astrocytes did not form a syncytium. Surface chemistry appeared to play only a minor role in the tissue reaction, suggesting that other design variables, such as porosity, stiffness, or implant size/surface area may play a larger role. We also found that the implantation of penetrating brain implants was sufficient to decrease hippocampal neurogenesis, regardless of implant location, in all cases but lattice implants with reduced surface area/stiffness. This suggests that the tissue response may be passively antagonized by design variables built into future CNS devices. |
