(136c) Rational Design and Interspecies Translation of Glucose-Responsive Insulins and Insulin Delivery Systems By Computational Modeling

Glucose-responsive insulin (GRI) is a type of ‘smart’ insulin therapy for diabetic patients to mitigate the risk of the tight therapeutic range of blood glucose concentration in typical insulin replacement regimens. The roles of GRI are to prevent the hypoglycemic risk of diabetic patients from exogenous insulin treatment and improve patients’ lives by avoiding regular subcutaneous insulin injections, which are major potential sources of user error. Researchers have investigated various novel approaches to realize GRIs for decades. These approaches can be classified into three categories: intrinsic GRIs, glucose-responsive particles, and glucose-responsive insulin delivery devices. Intrinsic GRIs are modified insulin analogues that exhibit inherent glucose-responsiveness. Glucose-responsive particles are freely circulating micro- or nanoparticles that release insulin during hyperglycemia. Finally, glucose-responsive insulin delivery devices are transdermal insulin delivery devices only activated during hyperglycemia. We developed a computational model named PAMERAH (Pharmacokinetic Algorithm Mapping GRI Efficacies in Rodents and Humans), which enables the assessment of GRI therapeutic efficacy and translatability between humans, rats, and mice. These models provide an important tool for the evaluation of clinical trial outcomes. PAMERAH is constructed of the compartmental physiological model of humans and rodents, with the additional GRI compartments are embedded. Our previous studies focused on intrinsic GRIs – here, we expand our model to other types of glucose-responsive systems: glucose-responsive particles and devices. The model examines efficacies after subcutaneous injection of GRIs over 24 hours, and we map the therapeutic range of GRIs based on dose and kinetic parameters. With the PAMERAH construct, we demonstrate that different classes of GRI exhibit different translatability patterns between humans and rodents, and these results show the capability of our model to simulate complicated therapeutic constructs. Hence, we envision PAMERAH as an effective tool for GRI development and a route toward realizing model guided clinical treatment for humans.