Background: The stability of insulin lispro for use in continuous subcutaneous insulin infusion (CSII) therapy was evaluated using a stress test incorporating high temperature and mechanical agitation combined with simulated basal/bolus administration.
Methods: Insulin lispro formulation contained in MiniMed 507c (Medtronic MiniMed, Sylmar, CA), H-TRONplus (Disetronic Medical Systems, St. Paul, MN), and D-TRON CSII (Disetronic Medtronic Systems) devices was subjected to a stress test involving exposure to elevated temperature (37 degrees C) and mechanical agitation (shaking at 100 strokes/min) for 7 days. CSII devices were programmed for continuous infusion at 0.8 U/h (19.2 U/day), and three 6-U bolus doses were also manually delivered each day (18 U/day). Formulation infused from the devices was collected every 24 h over the 7-day study. The material obtained from each 24-h period was analyzed to assess the physicochemical properties of insulin lispro and the overall formulation. Insulin lispro potency, purity, high-molecular-weight protein (HMWP), and m-cresol content were determined using high-performance liquid chromatographic methods. Solution pH, delivered volume, and formulation physical appearance were also evaluated.
Results: The pH of the infused formulation remained unchanged throughout the entire study. Insulin lispro potency for each device remained within 95-105% of label claim each day of the stress test, indicating that adsorption or surface-induced denaturation of insulin lispro did not occur. Levels of insulin lispro deamidated at position 21 in the A-chain were essentially constant during the 7-day stress test. A time-dependent increase in HMWP and other insulin lispro chemical transformation products was observed over the testing period for each device. However, the levels of these known degradation products remained well below published specifications for these analytical properties. The concentration of the antimicrobial agent, m-cresol, was decreased because of absorption, but the levels remained sufficient to provide protection against microbial contamination. Total delivered volume results were in good agreement with expected values and confirmed that formulation viscosity remained unchanged. Samples collected from the infusion sets were clear and free of precipitates. Occlusion alarms did not occur, and no other electronic or mechanical malfunctions of the CSII devices were encountered.
Conclusions: Insulin lispro demonstrates appropriate physicochemical stability for use in MiniMed 507c, H-TRONplus, and D-TRON CSII devices.