In silico evaluation of pramlintide dosing algorithms in artificial pancreas systems

Area of Science:

• Biomedical Engineering
• Endocrinology
• Computational Biology

Background:

• Pramlintide delays gastric emptying, making it a potential adjunct to insulin in artificial pancreas (AP) systems.
• Limited availability of pramlintide simulation models hinders in silico testing of combined therapies for type 1 diabetes (T1DM).

Purpose of the Study:

• To integrate a pramlintide pharmacokinetics/pharmacodynamics model into the T1DM UVA/Padova simulator.
• To adjust and validate four insulin-plus-pramlintide control algorithms for AP systems.
• To evaluate the performance of pramlintide-augmented AP strategies compared to insulin-alone controllers.

Main Methods:

• Incorporation of a recent pramlintide PK/PD model into the T1DM UVA/Padova simulator.
• Development and adjustment of four control algorithms: insulin-only and insulin-plus-pramlintide (bolus or ratio-based).
• Comparative analysis of simulated glycemic control metrics, focusing on time in range.

Main Results:

• Insulin-plus-pramlintide algorithms demonstrated improved time in range (3.00%–10.53%) compared to insulin-alone strategies.
• The simulation results align with findings from existing clinical trials.
• The study successfully validated the pramlintide model within the AP simulator.

Conclusions:

• Simulated insulin-plus-pramlintide strategies offer enhanced glycemic control in T1DM management.
• The developed simulation framework facilitates in silico testing of novel AP control algorithms.
• Future research should focus on patient-specific model personalization and testing under diverse conditions.