In-vitro and in-vivo performance studies of a porous infusion catheter designed for intraparenchymal delivery of therapeutic agents of varying size

Abstract

BackgroundLimitations have previously existed for the use of brain infusion catheters with extended delivery port designs to achieve larger distribution volumes using convection-enhanced delivery (CED), due to poor transmittance of materials and uncontrolled backflow. The goal of this study was to evaluate a novel brain catheter that has been designed to allow for extended delivery and larger distribution volumes with limited backflow of fluid. It was characterized using a broad range of therapeutic pore sizes both for transmittance across the membranes to address possible occlusion and for distribution in short term infusion studies, both in-vitro in gels and in-vivo in canines. MethodsBrain catheters with pore sizes of 10, 12, 15, 20 and 30 µm were evaluated using three infusates prepared in 0.9% sterile saline with diameters approximating 2, 5, and 30 nm, respectively. Magnevist™ was chosen as the small molecule infusate to mimic low-molecular weight therapeutics. Galbumin™ served as a surrogate for an assortment of proteins used for brain cancer and Parkinson’s disease. Gadoluminate™ was used to assess the distribution of large therapeutics, such as adeno-associated viral particles and synthetic nanoparticles. The transmittance of the medium and large tracer particles through catheters of different pore size (15, 20 and 30 µm) was measured by MRI and compared with the measured concentration of the control. Infusions into 0.2% agarose gels were performed in order to evaluate differences in transmittance and distribution of the small, medium, and large tracer particles through catheters with different pore sizes (10, 12, 15, 20 and 30 µm). In-vivo infusions were performed in the canine in order to evaluate the ability of the catheter to infuse the small, medium, and large tracer particles into brain parenchyma at high flow rates through catheters with different pore sizes (10, 15, and 20 µm). Two catheters were stereotactically inserted into the brain for infusion, one per hemisphere, in each animal (N = 6). ResultsThe transmittance of Galbumin and Gadoluminate across the catheter membrane surface was 100% to within the accuracy of the measurements. There was no evidence of any blockage or retardation of any of the infusates. Catheter pore size did not appear to significantly affect transmittance or distribution in gels of any of the molecule sizes in the range of catheter pore sizes tested. There were differences in the distributions between the different tracer molecules: Magnevist produced relatively large distributions, followed by Gadoluminate and Galbumin. We observed no instances of uncontrolled backflow in a total of 12 in-vivo infusions. In addition, several of the infusions resulted in substantial amounts remaining in tissue. We expect the in-tissue distributions to be substantially improved in the larger human brain. Comparison with existing methodsThe new porous brain catheter performed well in terms of both backflow and intraparenchymal infusion of molecules of varying size in the canine brain under CED flow conditions. ConclusionsOverall, the data presented in this report support that the novel porous brain catheter can deliver therapeutics of varying sizes at high infusion rates in the brain parenchyma, and resist backflow that can compromise the efficacy of CED therapy. Additional work is needed to further characterize the brain catheter, including animal toxicity studies of chronically implanted brain catheters to lay the foundation for its use in the clinic.