Simulation and Optimisation of a Self-regulating Insulin Delivery System

The delivery of insulin is an integral part of the treatment of diabetes. It has been shown that an implantable polymeric system which releases insulin in response to blood glucose levels is feasible. This work aims to guide further experimental development of this system by constructing a mathematical model of the polymer matrix and analysing its functional characteristics by computer simulations. The system is an implantable polymer containing tri-lysyl insulin and the enzyme glucose-oxidase, and the feedback mechanism is based on the enzymatic reaction between glucose and glucose-oxidase. Acid produced from this reaction reduces the pH in the microenvironment of the polymer, which causes an increase in insulin solubility and release rate. The model was developed on the basis of the physical and chemical properties of the system, which were chosen in the light of direct observations with scanning electron microscopy combined with experimental measurements and reported values, and was validated by comparison with experimental results and by verification of some of its assumptions. Optimisation was undertaken by simulating the effect of different parameters of the system on its performance. Enzyme concentration, pore length, particle size and insulin loading were found to have surprisingly little effect. However, performance was significantly improved by using a hypothetical insulin molecule with a different solubility characteristic. The study can therefore provide a rational basis for the experimental development of a polymer-based artificial pancreas.