Search

Convection-enhanced delivery (CED) is a bulk flow-driven process. Its properties permit direct, homogeneous, targeted perfusion of CNS regions with putative therapeutics while bypassing the blood-brain barrier. Development of surrogate imaging tracers that are co-infused during drug delivery now permit accurate, noninvasive real-time tracking of convective infusate flow in nervous system tissues. The potential advantages of CED in the CNS over other currently available drug delivery techniques, including systemic delivery, intrathecal and/or intraventricular distribution, and polymer implantation, have led to its application in research studies and clinical trials. The authors review the biophysical principles of convective flow and the technology, properties, and clinical applications of convective delivery in the CNS.

Direct interstitial infusion is a technique capable of delivering agents over both small and large dimensions of brain tissue. However, at a sufficiently high volumetric inflow rate, backflow along the catheter shaft may occur and compromise delivery. A scaling relationship for the finite backflow distance along this catheter in pure gray matter (x(m)) has been determined from a mathematical model based on Stokes flow, Darcy flow in porous media, and elastic deformation of the brain tissue: x(m) = constant Q(o)(3)R(4)r(c)(4)G(-3)mu(-1) 1/5 [corrected] = volumetric inflow rate, R = tissue hydraulic resistance, r(c) = catheter radius, G = shear modulus, and mu = viscosity). This implies that backflow is minimized by the use of small diameter catheters and that a fixed (minimal) backflow distance may be maintained by offsetting an increase in flow rate with a similar decrease in catheter radius. Generally, backflow is avoided in rat gray matter with a 32-gauge catheter operating below 0.5 microliter/min. An extension of the scaling relationship to include brain size in the resistance term leads to the finding that absolute backflow distance obtained with a given catheter and inflow rate is weakly affected by the depth of catheter tip placement and, thus, brain size. Finally, an extension of the model to describe catheter passage through a white matter layer before terminating in the gray has been shown to account for observed percentages of albumin in the corpus callosum after a 4-microliter infusion of the compound to rat striatum over a range of volumetric inflow rates.

Object: Selective treatment of central nervous system (CNS) structures holds therapeutic promise for many neurological disorders, including Parkinson's disease (PD). The ability to inhibit or augment specific neuronal populations within the CNS reliably by using present therapeutic techniques is limited. To overcome this problem, the authors modeled and developed a method in which convection was used to deliver compounds to deep brain nuclei in a reproducible, homogeneous, and targeted manner. To determine the feasibility and clinical efficacy of convective drug delivery for treatment of a neurological disorder, the investigators selectively ablated globus pallidus internus (GPi) neurons with quinolinic acid (QA), an excitotoxin, in the 1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine (MPTP)-induced model of primate parkinsonism., Methods: After the parameters of convective distribution to the GPi were confirmed by infusion of biotinylated albumin into the GPi of a primate (Macaca mulatta), seven adult monkeys of this species were rendered either fully parkinsonian by intravenous injections of MPTP (five animals) or hemiparkinsonian by a right-sided intracarotid injection of this agent (two monkeys). Using convection-enhanced delivery to the GPi, animals were infused with either QA (three fully parkinsonian, two hemiparkinsonian) or saline (two fully parkinsonian). The three fully parkinsonian animals that underwent GPi lesioning with QA had substantial improvement of PD symptoms, manifested by a marked increase in activity (34 +/- 2.5%; mean +/- standard deviation) and dramatic improvement of parkinsonian clinical scores. In contrast, the control animals did not improve (activity monitor change = -1.5 +/- 0.5%). The two hemiparkinsonian animals that underwent QA lesioning of the GPi had dramatic recovery of extremity use. Histological examination revealed selective neural ablation of GPi neurons (mean loss 87%) with sparing of surrounding gray and white matter structures. No animal developed worsening signs of PD or neurological deficits after infusion., Conclusions: Convection-enhanced delivery of QA permits selective, region-specific (GPi), and safe lesioning of neuronal subpopulations, resulting in dramatic improvement in parkinsonian symptomatology. The properties of convection-enhanced delivery indicate that this method could be used for chemical neurosurgery for medically refractory PD and that it may be ideal for cell-specific therapeutic ablation or trophic treatment of other targeted structures associated with CNS disorders.

Object: Although recent studies have shown that convection can be used to distribute macromolecules within the central nervous system (CNS) in a homogeneous, targeted fashion over clinically significant volumes and that the volume of infusion and target location (gray as opposed to white matter) influence distribution, little is known about other factors that may influence optimum use of convection-enhanced distribution. To understand the variables that affect convective delivery more fully, we examined the rate of infusion, delivery cannula size, concentration of infusate, and preinfusion sealing time., Methods: The authors used convection to deliver 4 microl of 14C-albumin to the striatum of 40 rats. The effect of the rate of infusion (0.1, 0.5, 1, and 5 microl/minute), cannula size (32, 28, and 22 gauge), concentration of infusate (100%, 50%, and 25%), and preinfusion sealing time (0 and 70 minutes) on convective delivery was examined using quantitative autoradiography, National Institutes of Health image analysis software, scintillation analysis, and histological analysis. Higher rates of infusion (1 and 5 microl/minute) caused significantly (p < 0.05) more leakback of infusate (22.7+/-11.7% and 30.3+/-7.8% [mean+/-standard deviation], respectively) compared with lower rates (0.1 microl/minute [4+/-3.6%] and 0.5 microl/minute [5.2+/-3.6%]). Recovery of infusate was significantly (p < 0.05) higher at the infusion rate of 0.1 microl/minute (95.1+/-2.8%) compared with higher rates (85.2+/-4%). The use of large cannulae (28 and 22 gauge) produced significantly (p < 0.05) more leakback (35.7+/-8.1% and 21.1+/-7.5%, respectively) than the smaller cannula (32 gauge [5.2+/-3.6%]). Varying the concentration of the infusate and the preinfusion sealing time did not alter the volume of distribution, regional distribution, or infusate recovery., Conclusions: Rate of infusion and cannula size can significantly affect convective distribution of molecules, whereas preinfusion sealing time and variations in infusate concentration have no effect in this small animal model. Understanding the parameters that influence convective delivery within the CNS can be used to enhance delivery of potentially therapeutic agents in an experimental setting and to indicate the variables that will need to be considered for optimum use of this approach for drug delivery in the clinical setting.

Object: Although many macromolecules have treatment potential for peripheral nerve disease, clinical use of these agents has been restricted because of limitations of delivery including systemic toxicity, heterogeneous dispersion, and inadequate distribution. In an effort to overcome these obstacles, the authors examined the use of convection to deliver and distribute macromolecules into peripheral nerves., Methods: For convective delivery, the authors used a gas-tight, noncompliant system that provided continuous flow through a small silica cannula (inner diameter 100 microm, outer diameter 170 microm) inserted into a peripheral nerve. Increases in the volume of infusion (Vi) (10, 20, 30, 40, and 80 microl) of 14C-labeled (nine nerves) or gadolinium-labeled (two nerves) albumin were infused unilaterally or bilaterally into the tibial nerves of six primates (Macaca mulatta) at 0.5 microl/minute. The volume of distribution (Vd), percentage recovery, and delivery homogeneity were determined using quantitative autoradiography, an imaging program developed by the National Institutes of Health, magnetic resonance (MR) imaging, scintillation counting, and kurtosis (K) analysis. One animal that was infused bilaterally with gadolinium-bound albumin (40 microl to each nerve) underwent MR imaging and was observed for 16 weeks after infusion. The Vd increased with the Vi in a logarithmic fashion. The mean Vd/Vi ratio over all Vi was 3.7+/-0.8 (mean+/-standard deviation). The concentration across the perfused region was homogeneous (K=-1.07). The infusate, which was limited circumferentially by the epineurium, followed the parallel arrangement of axonal fibers and filled long segments of nerve (up to 6.8 cm). Recovery of radioactivity was 75.8+/-9%. No neurological deficits arose from infusion., Conclusions: Convective delivery of macromolecules to peripheral nerves is safe and reliable. It overcomes obstacles associated with current delivery methods and allows selective regional delivery of putative therapeutic agents to long sections of nerve. This technique should permit the development of new treatments for numerous types of peripheral nerve lesions.

Object: Because of the limited penetration of macromolecules across the blood-spinal cord barrier, numerous therapeutic compounds with potential for treating spinal cord disorders cannot be used effectively. The authors have developed a technique to deliver and distribute macromolecules regionally in the spinal cord by using convection in the interstitial space., Methods: The authors designed a delivery system connected to a "floating" silica cannula (inner diameter 100 microm, outer diameter 170 microm) that provides for constant volumetric inflow to the spinal cord. A solution containing albumin that was either unlabeled or labeled with carbon-14 or gadolinium was infused at various volumes (3, 6, 10, 20, 40, or 50 microl) at a rate of 0.1 microl/minute into the spinal cord dorsal columns of nine swine and into the lateral columns of three primates (Macaca mulatta). Volume of distribution (Vd), concentration homogeneity, and percentage of recovery were determined using scintillation analysis, kurtosis calculation (K), and quantitative autoradiography (six swine), magnetic resonance imaging (one swine and three primates), and histological analysis (all animals). Neurological function was observed for up to 3 days in four of the swine and up to 16 weeks in the three primates. The Vd of 14C-albumin was linearly proportional (R2=0.97) to the volume of infusion (Vi) (Vd/Vi=4.4+/-0.5; [mean+/-standard deviation). The increases in Vd resulting from increases in Vi were primarily in the longitudinal dimension (R2=0.83 in swine; R2=0.98 in primates), allowing large segments of spinal cord (up to 4.3 cm; Vi 50 microl) to be perfused with the macromolecule. The concentration across the area of distribution was homogeneous (K=-1.1). The mean recovery of infused albumin from the spinal cord was 85.5+/-5.6%. Magnetic resonance imaging and histological analysis combined with quantitative autoradiography revealed the albumin infusate to be preferentially distributed along the white matter tracts. No animal exhibited a neurological deficit as a result of the infusion., Conclusions: Regional convective delivery provides reproducible, safe, region-specific, and homogeneous distribution of macromolecules over large longitudinal segments of the spinal cord. This delivery method overcomes many of the obstacles associated with current delivery techniques and provides for research into new treatments of various conditions of the spinal cord.

Many novel experimental therapeutic agents, such as neurotrophic factors, enzymes, biological modifiers, and genetic vectors, do not readily cross the blood-brain barrier. An effective strategy to deliver these compounds to the central nervous system is required for their application in vivo. Under normal physiological conditions, brain interstitial fluid moves by both bulk flow (convection) and diffusion. It has recently been shown that interstitial infusion into the white matter can be used to increase bulk flow, produce interstitial convection, and efficiently and homogeneously deliver drugs to large regions of brain without significant functional or structural damage. In theory, even more uniform distribution is likely in gray matter. In the current study, four experiments were performed to examine if convection-enhanced delivery could be used to achieve regional distribution of large molecules in gray matter. First, the volume and consistency of anatomical distribution of 20 microliters of phaseolus vulgaris-leukoagglutinin (PHA-L; molecular weight (MW) 126 kD) after continuous high-flow microinfusion into the striatum of five rats over 200 minutes were determined using immunocytochemistry and quantified with image analysis. Second, the concentration profile of 14C-albumin (MW 69 kD) infused under identical conditions was determined in four hemispheres using quantitative autoradiography. Third, the volume of distribution after convection-enhanced infusion of 250 or 500 microliters biotinylated dextran (b-dextran, MW 10 kD), delivered over 310 minutes into the caudate and putamen of a rhesus monkey from one (250 microliters) or two (500 microliters) cannulas, was determined using immunocytochemistry and quantified with image analysis. Finally, the ability to target all dopaminergic neurons of the nigrostriatal tract via perfusion of the striatum with subsequent retrograde transport was assessed in three experiments by immunohistochemical analysis of the mesencephalon following a 300-minute infusion of 27 microliters horseradish peroxidase-labeled wheat germ agglutinin (WGA-HRP) into the striatum. Convection-enhanced delivery reproducibly distributed the large-compound PHA-L throughout the rat striatum (the percent volume of the striatum perfused, Vs, was 86% +/- 5%; mean +/- standard deviation) and produced a homogeneous tissue concentration in the perfused region (concentration of 14C-albumin relative to infusate concentration 30% +/- 5%). In the monkey, the infusion widely distributed b-dextran within the striatum using one cannula (caudate and putamen Vs = 76% and 76%) or two cannulas (Vs = 90% and 71%).(ABSTRACT TRUNCATED AT 400 WORDS)

High-flow microinfusion provides a means for delivering macromolecules to large volumes of brain in easily obtainable time intervals. Slowly degraded approximately 180-kDa macromolecules, delivered at a constant volumetric flow rate of 3 microliters/min into homogeneous brain tissue (e.g., gray matter), would penetrate to a 1.5-cm radius in 12 h. The predicted concentration profile is relatively flat until it declines precipitously at the flow front. Hence, tissues are dosed rather uniformly, providing control over the undesired toxicity that may occur with alternative methods that depend on large concentration gradients for tissue transport. The penetration advantage of high-flow (convective) over low-flow (diffusive) microinfusion has been assessed at fixed pharmacodynamic effect. A 12-h high-flow microinfusion of a macromolecule degraded with a characteristic time of 33.5 h would provide 5- to 10-fold increases in volume over low-flow infusion and total treatment volumes > 10 cm3. Slower degradation rates would result in larger treatment volumes; more rapid degradation rates would reduce the volume but still favor convective over diffusive administration. This technique may be applicable to a variety of diagnostic and therapeutic agents such as radioimmunoconjugates, immunotoxins, enzymes, growth factors, and oligonucleotides.