Despite aggressive surgery and post-operative radiation and chemotherapy, the prognosis is poor for glioblastoma patients. Anti-angiogenic therapy with compounds such as endostatin could delay the onset of relapse. However, the short systemic half-life of this proteins as well as the blood-brain barrier makes the use of this therapy difficult for brain cancer patients. The aim of this project is to develop and implant genetically engineered producer cells secreting endostatin that are encapsulated in calcium cross-linked alginate gel beads. Encapsulation of cells within alginate gels has a potential as a sustained release system in addition to the fact that the encapsulation technology protects the cells from rejection by the immune system. Human embryonal kidney 293 cells have been transfected with the gene for endostatin. These cells have been encapsulated in calcium cross-linked alginate gels and optimized for the secretion of endostatin. Alginate gel beads implanted into rat brain have shown only a moderate loss in cell viability but extended endostatin release for periods of up to 12 months. Visualization of the anti-angiogenic effect on C6 rat glioma growth, tumor vasculature and microhemodynamics has been demonstrated by using intravital video microscopy. The data indicates that endostatin greatly affects tumor-associated microcirculation but does not appear to affect normal microcirculation. The local delivery of endostatin seems to specifically affect tumor-associated microvessels by reduction of the vessel density, diameter and functionality. Tumor cell migration and invasion was greatly reduced in the endostatin treated animals