A new C‐peptide correction model used to assess bioavailability of regular human insulin

Abstract

The clinical assessment of new formulations of human insulin is problematic due to the inability to distinguish between endogenous insulin and exogenously administered insulin. The usual methods to surmount the problem of distinguishing between endogenous and exogenous human insulin include evaluation in subjects with no or little endogenous insulin, hyper-insulinemic clamp studies or the administration of somatostatin to suppress endogenous insulin secretion. All of these methods have significant drawbacks. This paper describes a method for C-Peptide correction based upon a mixed effects linear regression of multiple time point sampling of C-Peptide and insulin. This model was able to describe each individual's insulin to C-Peptide relationship using the data from four different phase I clinical trials involving both subjects with and without type 2 diabetes in which insulin and C-Peptide were measured. These studies used hyper-insulinemic euglycemic clamps or meal challenges and subjects received insulin or Glucagon-like peptide 1 (GLP-1). It was possible to determine the exogenously administered insulin concentration from the measured total insulin concentration. A simple statistical technique can be used to determine each individual's insulin to C-Peptide relationship to estimate exogenous and endogenous insulin following the administration of regular human insulin. This technique will simplify the assessment of new formulations of human insulin.