T. Yu,1 M. Sinha,2 M. Hillard,2 C. Sherwin,1 S. Russell,2 M. Spigarelli1; 1University of Utah, Salt Lake City, UT, 2Massachusetts General Hospital, Boston, MA

BACKGROUND: Slow absorption of insulin, even for “rapid-acting” insulin analogs, is a major challenge for effective diabetes control. This study aimed to assess the population pharmacokinetics (PK) of insulin delivered through intradermal and subcutaneous routes to determine the preferred route of insulin delivery for rapid absorption.
METHODS: Patients aged ≥ 18 years with type 1 diabetes for ≥ 1 year were recruited for this crossover study. Each patient received 5 units of insulin lispro (Humalog, Eli Lilly) via intradermal (MicronJet, Nanopass) or subcutaneous injection. Blood samples were collected at intervals of 20 minutes or less over 5 hours. Insulin concentrations were measured by quantitative chemiluminescent immunoassay. Nonlinear mixed effects modeling (NONMEM 7.2) were used to develop population PK models.
RESULTS: A total of 15 patients were recruited with median (range) age of 36 (22-62) years and anti-insulin antibody titer 0.065 (0-0.33) nmol/L (normal range < 0.02 nmol/L). The insulin PK after intradermal and subcutaneous injection was described well by a two-compartment model with lag time. Intradermal delivery resulted in shortened Tmax (0.61 vs. 0.83 hr, p < 0.05), increased Vcentr (609 vs. 25 L, p < 0.001) and increased Ka (4.91 vs. 0.762 hr-1, p < 0.001) relative to subcutaneous delivery. Injection bleb formation significantly reduced insulin Vcentr (p < 0.001) and anti-insulin antibody improved model fit of insulin PK (p > 0.05) after intradermal injection. Anti-insulin antibody significantly decreased insulin CLcentr (p < 0.01) and Vcentr (p < 0.001) after subcutaneous injection.
CONCLUSION: The results demonstrated the feasibility of utilizing intradermal delivery to increase the speed of insulin absorption.