Wirelessly controlled, bioresorbable drug delivery device with active valves that exploit electrochemically triggered crevice corrosion

Abstract

Implantable drug release platforms that offer wirelessly programmable control over pharmacokinetics have potential in advanced treatment protocols for hormone imbalances, malignant cancers, diabetic conditions, and others. We present a system with this type of functionality in which the constituent materials undergo complete bioresorption to eliminate device load from the patient after completing the final stage of the release process. Here, bioresorbable polyanhydride reservoirs store drugs in defined reservoirs without leakage until wirelessly triggered valve structures open to allow release. These valves operate through an electrochemical mechanism of geometrically accelerated corrosion induced by passage of electrical current from a wireless, bioresorbable power-harvesting unit. Evaluations in cell cultures demonstrate the efficacy of this technology for the treatment of cancerous tissues by release of the drug doxorubicin. Complete in vivo studies of platforms with multiple, independently controlled release events in live-animal models illustrate capabilities for control of blood glucose levels by timed delivery of insulin.

Fig. 1. Wirelessly programmable, bioresorbable drug delivery system.

(A) Illustration and images of an implantable and wirelessly controlled bioresorbable drug delivery system with an electrical triggering unit that includes a radio frequency (RF) power harvester with a Mg coil, a silicon nanomembrane (Si NM) diode, and a Mg/SiO₂/Mg capacitor, [A(i)]. (B) RF behavior (scattering parameter, S11) of the harvester (black, experiment; blue, simulation). The resonance frequency is ~5 MHz, selected to allow magnetic coupling with little parasitic absorption by biological tissues. (C) Simulated inductance (black) and Q factor (red) of a single coil with a diameter of 16 mm. (D) Transmitting power as a function of the distance between the transmitting coil and the device (black, experiments; red, simulation). Experimental data are means ± SD; n = 3. (E) Images of wirelessly controlled release of a blue dye in water during immersion in phosphate-buffered solution (PBS; pH 7.4). (F) Accelerated dissolution of an entire system due to immersion in PBS (pH 7.4) at an elevated temperature, 85°C. Photo credit: Jahyun Koo, Korea University.

Fig. 2. Geometrically accelerated corrosion as the mechanism for opening a gate.

(A) Design of an implantable and wirelessly controlled, bioresorbable drug delivery system. (B) Illustration and image of a Mg gate defined by an opening through a cover of PBTPA. The locations where the edges of this opening meet the underlying Mg create regions of crevice corrosion. The insets show optical microscope (OM; left) and scanning electron microscope (SEM; right) images of the electrochemically accelerated corrosion that occurs along these edge locations. (C) Back side view images of an electrochemically etched Mg gate (3 mm by 3 mm) at an applied potential of 1 V (~5 mA) for 0, 30, 90, and 180 s, respectively. (D) Nyquist plot: Z′, real part of impedance; Z′′, imaginary part of impedance. (E) OM image of the electrochemically accelerated corrosion of a Mo gate (3 mm by 3 mm, 10 μm thick) at an applied voltage of 5 V (~6 mA) for 60 s. (F) Three-dimensional (3D) topography image and surface profile measurement of the corroded Mo gate near the edge of the PBTPA (1:1:2.5) opening, obtained using a 3D optical profiler (Nexview, Zygo Corporation, USA). Photo credit: Ji-Hyeon Park, Northwestern University.

Fig. 3. Trigger and release behaviors of bioresorbable drug delivery vehicles with electrochemical control.

(A) Dependence of the trigger time on applied potential at fixed Mg gate size (30 μm thickness, 3 mm by 3 mm area). (B) Effect of the exposed area of the gate on trigger time at a fixed applied potential (0.5 V) and gate thickness (30 μm). (C) Dependence of trigger time on thickness of the gate, for a fixed applied potential (0.5 V) and gate area (3 mm by 3 mm). (D) Comparison of the trigger time of Mg (30 μm at 0.2, 0.5, and 1 V) and Mo (10 μm at 1 V) gates for various perimeters of the gates. (E) Kinetics of the release of doxorubicin after triggering to open an area of 2 mm by 2 mm in the Mg gates. (F) Leakage of drug (doxorubicin) through the layers of material used for the reservoir and the gates [black, Mg (50 μm thick); red, Mo (10 μm thick); and blue, PBTPA 1:4:7 (100 μm thick)] during immersion in PBS at body temperature (37°C). Photo credit: Jahyun Koo, Korea University.

Fig. 4. In vitro evaluation of NIH-3T3 cell and tumor cell (HeLa, HepG2, and MDA-MB-231) viability after triggered release of drug (doxorubicin).

(A to C) Series of optical and fluorescence images of NIH-3T3 cells with phalloidin staining of cell F-actin (green; ii) and DAPI staining of the nuclei (blue; iii) before and after triggered release. The inset shows an image of a test platform used to study cell viability. (A and B) Images of the initial cell configuration and after 1-hour incubation with a device immersed in Dulbecco’s modified Eagle’s medium (DMEM) before triggering, respectively. (C) Images after triggering and release of doxorubicin from the device. Relative viability of (D) NIH-3T3, (E) HeLa, (F) HepG2, and (G) MDA-MB-231 cells before and after triggering (release point, 1 hour). Data are means ± SD; n = 3 per group. Statistica software (Version 6.0) was used for statistical analysis followed by a t test. *P < 0.05 compared to any other group. Photo credit: Hojun Kim, University of Illinois at Urbana-Champaign.

Fig. 5. In vivo operation of a wirelessly programmable, bioresorbable drug release vehicle with three separately addressable drug reservoirs for regulation of blood glucose.

(A) Illustration of a system with three separate reservoirs and wireless stimulator units. (B) Simulated and (C) experimentally measured (data are means ± SD; n = 3 per group) voltages of the three separate harvesters at different operating frequencies, to demonstrate minimal cross-talk. (D and E) Images of a device and surgical procedures for implantation in the subdermal area of the back of a rat model. (F) Changes in blood glucose level induced by wirelessly triggering the release of insulin sequentially from the three reservoirs. RF with different frequencies (5, 10, and 15 MHz) triggered each reservoir. After each release event and a subsequent hour of monitoring, glucose was injected under the skin to return the glucose level to its initial value within 30 min. Photo credit: Jahyun Koo, Korea University.

Fig. 6. In vivo demonstration of a wirelessly programmable, bioresorbable drug release vehicle with three independent reservoirs and a cuff structure as a local anesthetic for mitigation of pain.

(A) Illustration of a system with triple reservoirs and a cuff structure as an interface to the sciatic nerve of a rat model. The inset highlights the design details and the nerve interface. (B and C) Schematic illustration and image of the device. (D) Images at various stages of implantation of the device and integration of its flexible poly(lactic-co-glycolic acid) (PLGA) cuff (4 mm by 8 mm, ~20 μm thick) with the sciatic nerve to hold the position of the reservoir and to localize the release to the nerve. (E) Changes in the maximum EMG signal associated with wirelessly triggered released of lidocaine from each of the three reservoirs in series. Here, RF with different frequencies (5, 10, and 15 MHz) triggered each reservoir. After each release event of lidocaine (0.2%, ~20 μl), the maximum EMG was recorded from the tibialis anterior (TA) muscle (data are means ± SD; n = 2). (F) Representative electromyogram (EMG) data for every 15 min after the trigger events and corresponding release of lidocaine (0.2%, ~20 μl) in series (black, baseline; red, 15 min after the first trigger; blue, 15 min after the second trigger; and purple, 15 min after the third trigger). Photo credit: Jahyun Koo, Korea University.

Fig. 7. In vivo biocompatibility studies.

Hematoxylin and eosin images of stained tissue sections at 5 weeks after implantation of (A) devices and (B) samples of HDPE. Middle and right images show adjacent muscle fascia and muscle tissue, respectively. (C) Evaluations of histological scoring for reference controls (HDPE) and test groups (Mg metal gate and Mo metal gate devices) (n = 3 per group). PMN, polymorphonuclear. Boxplots show the median (center line), the third and first quartiles (upper and lower edge of box, respectively), and the largest and smallest value that is ≤1.5 times the interquartile range (limits of upper and lower whiskers, respectively). Photo credit: Matthew MacEwan, Washington University in St. Louis.