Alzheimer’s disease (AD) is the most prevalent form of dementia in the elderly. As populations are aging in most parts of the world, the number of people with Alzheimer’s disease is expected to dramatically increase. This devastating neurodegenerative disorder remains the only major source of mortality without any effective disease-modifying treatment. Nevertheless, the progress made in understanding underlying disease mechanisms linked to amyloid and tau pathologies, led to the development of novel therapeutic strategies. The most clinically advanced experimental treatment is based on monoclonal antibodies (passive immunization) directed against the amyloid β peptide, a critical contributor to the neuropathology of Alzheimer’s disease. Although immunotherapy is able to clear cerebral amyloid deposits in patients, it has so far failed to slow down the loss of cognitive functions. In order to potentiate therapeutic effects, it may be critical to initiate the treatment before the first cognitive symptoms, and therefore chronically administer antibodies over years, possibly decades. This prompts the development of alternative strategies for the long-term delivery of therapeutic antibodies. To address this question, we have investigated an innovative approach based on a biologically active subcutaneous device encapsulating cells genetically engineered for the production of recombinant antibodies. In order to demonstrate that encapsulated cell technology can achieve peripheral continuous delivery of antibodies and lead to significant antibody exposure inside the brain, we rationally optimized parameters that are crucial to the rate of antibody secretion from such a cellular implant. In the first part of the thesis, we explored the utilization of lentiviral vectors for the genetic engineering of antibody secreting cell lines. We developed a dual vector strategy to significantly improve the efficiency of stable cell line generation. Moreover, the lentiviral transduction of C2C12 murine myoblasts allowed for the isolation of individual clones demonstrating high cell specific productivity, thus validating the use of C2C12 as a preclinical surrogate cell line for allogeneic implantation in disease mouse models. In the second part of the present work, we engineered a novel cell encapsulation device specially designed for subcutaneous implantation.We implemented a highly reproducible and scalable process to assemble devices supporting the implantation of large quantities of cells. Moreover, the in vivo optimization of initial cell seeding parameters enabled encapsulated cells to survive for months at high density inside the device. The final part of the present work assessed therapeutic efficacy of the newly developed cell encapsulation system for peripheral immunization against brain amyloid pathology. Devices secreting anti-Aβ antibodies were subcutaneously implanted in two transgenic mouse models of Alzheimer’s disease. Antibody levels in the mouse plasma reached 40 μg/ ml for several months, demonstrating long-term implant efficacy. Furthermore, we could demonstrate immunodecoration of amyloid plaques inside the brain with the delivered antibody, demonstrating penetration into the central nervous system. Remarkably, in the two mouse models of Alzheimer’s disease, brain amyloidosis was significantly reduced in the treated animals. Overall, this thesis work shows that peripheral implantation of encapsulated antibody-secreting myoblast cells is an effective strategy for passive immunization, particularly in the challenging context of delivering therapeutic amounts of antibodies to the central nervous system. Indeed, we demonstrate that anti-Aβ antibodies administered by subcutaneous bioactive implants can significantly decrease amyloid burden in mouse models of Alzheimer’s neuropathology.