(313g) On-Demand Drug Delivery Using Magnetic Thermoresponsive Membranes

INTRODUCTION: User-defined temporal control of flux through a membrane is of considerable interest for drug delivery, sensor technology, and biomolecule separation, among other potential applications. However, while many drug delivery technologies have been developed in order to achieve sustained drug release over an extended period of time, relatively few technologies currently exist for facilitating user-controlled or pulsatile release kinetics. This capability for effective "on-demand" drug delivery would be extremely beneficial for treating conditions requiring intermittent drug release (i.e. diabetes or chronic pain) or for treatments benefitting from highly flexible potential drug dosing regimens depending on the patient's response to a particular treatment. Radio frequency-activated implantable microchips containing drug-filled reservoirs can achieve repeated, on-demand pulsatile drug release if multiple reservoirs are present, but require the use of implanted electronics for proper function (1). Liposomes, microspheres, and microcapsules loaded with ferrofluids have also been developed which can be activated remotely by magnetic induction (2); however, in most cases, this heating results in only a single burst release event and cannot be applied to achieve reliable, repeatable drug release. Instead, an ideal device for on-demand drug release could safely contain a large concentration of a target drug and could be triggered repeatedly through a non-invasive mechanism to release drug as demanded by the patient (i.e. local pain relief) or as prescribed by a doctor (i.e. localized chemotherapy) over several months. To meet this need, we have developed a composite membrane-based device which consists of ethylcellulose (the membrane support), thermosensitive poly(N-isopropylacrylamide)-based nanogels (the switching entity) (3), and a superparamagnetic ferrofluid (the triggering entity) which achieves ?on-demand? and repeatable on-off switching of molecular flux upon the application of an oscillating magnetic field. The functionality of the membrane results from the localized induction heating of the ferrofluid nanoparticles within the nanogel-impregnated nanogel upon the application of an external oscillating magnetic field (4). Heat generated via magnetic induction is transferred to the adjacent thermosensitive nanogels, causing the nanogels to shrink and create void spaces in the membrane pores without significantly heating the surrounding tissues. This facilitates drug diffusion out of a membrane-based reservoir drug delivery device. When the magnetic field is turned off, the nanogels cool and re-swell, refilling the pores with gel and reducing the drug flux to a near-zero value. EXPERIMENTAL METHODS: Thermosensitive nanogels were synthesized by surfactant-free emulsion polymerization using 0.9g N-isopropylmethacrylamide, 0.5g N-isopropylacrylamide, 0.08g N,N-methylenebisacrylamide, 0.1g ammonium persulfate, and 150mL water. Nanogel suspensions in ethanol (6wt%) were then mixed with a 10wt% ethyl cellulose solution in ethanol and a polyethylene oxide-peptized ferrofluid to form the membranes by evaporation. The performance of the membranes was assayed in a controlled-temperature water bath by placing the membrane in a glass flow cell and tracking the flux of a sodium fluorescein solution into the receiving cell under different temperatures. Membrane-based reservoir devices were constructed by gluing two 1cm diameter membrane segments to the ends of a 3/8?OD silicone tube filled with sodium fluorescein. Magnetic activation experiments were then performed by placing the reservoir device into a solenoid coil operating at 260kHz and delivering magnetic field amplitudes from 0-20mT. Samples are collected at different time points and assayed for concentration using UV/VIS spectrophotometry. Membrane cytotoxicity was tested using the MTT assay. Membrane biocompatibility was assayed by implanting 1cm membrane disks subcutaneously in a rat, recovering the membrane after 2 months of implantation, and examining the surrounding tissues by histology. RESULTS: Controlled, on-off drug delivery could be achieved using the magnetic nanogel membranes triggered by either temperature or a magnetic field (see figure). Heating the membrane in the water bath (top figure) induced a ~20-fold increase in sodium fluorescein flux across a 25wt% nanogel/27wt% ferrofluid membrane. The observed flux at various temperatures in the transition region correlated directly with the hydrodynamic diameter of the nanogels in free suspension, permitting precise engineering of the drug flux at different temperatures based on the chemical composition of the nanogel used to prepare the membrane. In addition, the drug flux in the "on" state followed zero-order reservoir release kinetics, permitting adjustable dosing of drugs based on the duration of the "on" pulse. When membrane-capped reservoir devices are placed in an oscillating magnetic field, on-off release was remotely triggered by the magnetic field. Activation of the magnetic field induces local heating inside the membrane, resulting in a large increase in drug flux; de-activation of the magnetic field induces solution cooling and recovery of the initial, low-flux state. The devices release drug with zero-order kinetics over several days in the ?on? state, can be switched from ?on? to ?off? with only ~5-10 minute lag time on both heating and cooling, and remain functional over at least 10 thermal cycles (data not shown) to provide rapid, predictable, and repeatable long-term control over molecular flux. This approach has the advantage of being entirely materials-based in that non-invasive pulsatile release can be achieved without requiring implanted electrical components. Furthermore, the observed zero-order release kinetics from these devices in the ?on? state contrasts significantly with the release profiles typically observed in other pulsatile delivery devices which flood the body with a large drug dose upon activation. Cytotoxicity studies using the MTT assay indicated no significant toxicity related to either the membrane itself or the individual components of the membrane. When implanted subcutaneously in a rat, the devices induced only a minimal immune response, no significant inflammation, and maintained their triggerable flux properties after two months of implantation, with no significant reduction in either the maximum flux achievable or the difference between the low and high temperature flux levels. Thus, the membranes appear to have the potential for long-term functionality via in vivo implantation. CONCLUSION: Nanogel-loaded, magnetically-activated membranes can be used to achieve ?user-defined? control over drug release kinetics. The membranes can deliver controlled doses of drug at designed target temperatures and at desired times via non-invasive triggering of the devices in the presence of an oscillating magnetic field. Such devices may be useful for chemotherapy, chronic pain relief, and insulin delivery, among other applications. Acknowledgements: This research was funded by NIH grant GM073626 to DSK. TH acknowledges post-doctoral funding from the Natural Sciences and Engineering Research Council of Canada. REFERENCES (1) Santini, J.T.; Richards, A.C.; Scheidt, R.; Cima, M.J.; Langer, R. Angewandte Chemie Int. 2000, 39, 2396-2407. (2) Müller-Schulte, D. US Patent US20070148437A1. (3) Hoare, T.; Pelton, R. Macromolecules 2004, 37, 2544-2550. (4) Jackson, D.K.; Leeb, S.B.; Mitwalli, A.H.; Narvaez, P.; Fusco, D.; Lupton Jr., E. IEEE Trans. Indust. Elect. 1997, 44, 217-225.