In spite of the high therapeutic potential of many macromolecular therapeutic agents, transport issues often make it difficult to successfully apply them to neurodegenerative diseases. One barrier to systemic administration is low capillary permeability, i.e., the blood-brain barrier. To overcome this barrier, convection-enhanced delivery (CED) (direct infusion into tissue) supplements diffusion and increases the interstitial distribution of large molecules in the brain and spinal cord. Predictive models of distribution during CED would be useful in treatment optimization and planning. To account for large infusion volumes, such models should incorporate tissue boundaries and anisotropic tissue properties. We have previously developed three-dimensional transport models of CED for the spinal cord. Anisotropic tissue properties in white matter were assigned using tissue alignment data from diffusion tensor imaging (DTI). DTI provides non-destructive, three-dimensional visualization of brain and spinal cord structure with excellent soft tissue contrast and good spatial resolution (~0.00108 mm^3 in the rat spinal cord and ~0.00675 mm^3 in the rat brain). Within biological systems, water molecules undergo continuous Brownian motion, which results in translational diffusion. From measurements of MR diffusion-weighted images of water motion, information about fiber-tract trajectories within fibrous tissues and the structure of the underlying tissues can be calculated. In this study, white and grey matter tissue boundaries were determined using the diffusivity tensor of water in tissue measured by DTI. We have developed a methodology by which tissue properties determined on a voxel by voxel basis were assigned on an element by element basis to a computational (finite volume) transport model. The developed modeling approach is semi-automatic and utilizes porous media transport theory for rapid estimation of interstitial transport in heterogeneous tissue regions with anatomical boundaries in the brain and spinal cord. Validation studies in the dorsal horn of the rat spinal cord predict distributed volumes of albumin tracer that are comparable with previous experimental measures by Wood and coworkers for the small infusion volumes studied (<4 uL). Infusion into brain regions with more complex anatomical boundaries was also considered. Velocity fields for infusion into the corpus callosum found higher velocities confined primarily to the white matter tracts, consistent with previous infusion studies which show extensive spread in these regions. The developed model will also be used to determine tracer distributions in the vicinity of ventricles and the effect of endogenous flows on final distribution volumes.

Extended Abstract Status: File Uploaded