Pulsatile Drug Delivery
The Nuxoll Lab is developing an elegant materials-based approach to automated pulsatile drug delivery.  By stacking sacrificial barriers alternately with individual doses of drug, we build laminates which release the doses in discrete pulses.  Different drugs can be loaded into each layer and the barriers can be tuned to specific delay times, allowing, for instance, rapid assembly of personalized drug delivery chips from an assortment of pre-made polymer films.  Our long-term goal for this technology is the replacement of extended injection regimens (e.g., vaccination series or allergy treatments) with a single quickly-inserted polymer laminate.  There are at least two potential paths towards this goal, both of which involve carefully designed polymer composites to manipulate competing rate processes for controlled release.

Pused drug release from an erodible polymer laminate

Schematic demonstration of pulsed release from erodible polymer laminate. Upon immersion in water, the first solute pulse is released, followed by a period of no release while the top barrier layer erodes. Once the top barrier layer is gone, a second solute pulse is released, followed by erosion of the next barrier layer.  from J. Membr. Sci.

Encapsulating drug in an acid-sensitive hydrogel membrane, we can control the release of drug by controlling the flux of acid to the hydrogel.  Overlaying a composite hydrogel / nanoparticulate ZnO barrier membrane, we can block acid flux for a specific period of time based on the barrier thickness, ZnO loading, and ZnO position within the barrier.  We have demonstrated this control experimentally, tuning the delay time between pulses by an order of magnitude with slight changes in the barrier design.  We release different drugs alternately from the same device and have demonstrated devices with up to 10 programmed pulses (Gandhi, J. Membr. Sci., 2015).  We have demonstrated both delaminating and non-delaminating devices, with the latter being much more difficult to analytically design, prompting the development of a computational model to discern key design relationships (Gandhi, Chem. Eng. Sci., 2016).  We have also taken the first step toward adapting the system for physiological application by incorporating enzymes and demonstrating pulsatile release driven only by an external sink of glucose (Nishii, Chem. Eng. Res. Des., 2016).

BMPR devices

Configurations and release profiles for BMPR devices.  (Left) Configurations of two BMPR devices.  Solute-free depots are placed on top to guarantee initial adhesion to the impermeable wall.  Note that devices are not drawn to scale; actual aspect ratio resembles a coin.  (Right)  Solute release profiles for the devices at pH 7 (dotted) and pH 3 (solid).  No release observed at pH 7, while regular discrete pulses are observed at pH 3.  from J. Membrane. Sci.

Multiple pulse BMPR systemHow many pulses can one BMPR device provide?  (Left) Configuration of a 10-pulse BMPR device.  Note that devices are not drawn to scale; actual aspect ratio resembles a pair of coins.  (Right)  Solute release profiles for the devices at pH 7 and pH 3.  Most solutes, including the last four, released at their analytically predicted time, but the opportunity for wall-barrier contact problems increases with the number of doses.  from J. Membrane. Sci.

Degradable polymers
Pulsatile release using hydrolytically degradable polymer barriers is much simpler and has already been demonstrated in vivo.  The key shortcoming of this approach is the difficulty in predicting and scaling the timing of the release.  We have developed the first model of polymer degradation which can predict the erosion profile of polymers ranging from strongly bulk-eroding to strongly surface-eroding, based entirely on their degradation rate constant and their moisture diffusion coefficient.  This model includes integrated mass transport calculations, allowing the incorporation of autocatalysis and, for our end purposes, solute release from an underlying drug layer (Coffel, J. Membrane Sci.).

Transient polymer erosion profiles as a function of the Damköhler number, Φ, with corresponding pixel
maps demonstrating different erosion behaviors for strongly surface-eroding polymers (left) versus
strongly bulk-eroding polymers (right) versus an intermediate-type erosion behavior (middle).