Impact of Endogenous Insulin Adjustment Methods on the Assessment of Insulin Pharmacokinetics and Pharmacodynamics

Abstract

Assessing reliable insulin pharmacokinetics (PK) and pharmacodynamics (PD) following an exogenous insulin administration is a scientific challenge due to the circulating endogenous insulin. Effect of exogenous insulin on the endogenous insulin through the sensitive glucose homeostatic system in the body could further confound the PK/PD data. We compared results of PK/PD parameters of methods in literature for the estimation of insulin PK and PD, namely total method without adjustment (method 1), a simple baseline method (method 2), elimination rate constant method (method 3), empirical C-peptide correction method (method 4), and conventional PK/PD modeling method (method 5). We also developed new PK/PD modeling method reflecting the effect of glucose on endogenous insulin release based on C-peptide (method 6). Subject-level time-matching insulin and glucose concentrations without external glucose administration were collected from clinical trials with regular human insulin administration to healthy subjects under fasting conditions (n=23). Primary insulin PK parameters (i.e., the maximum concentration (Cmax) and area under the concentration-time profile (AUC)) were estimated using subjects’ insulin, glucose and C-peptide level data. Our results indicate that insulin AUC and Cmax differ by up to 32 and 17%, respectively, among different methods. Further, the contribution of endogenous insulin to total insulin AUC is estimated to be approximately 55% but varies depending upon estimation methods. Insulin PD parameters are apparently similar among different methods (e.g., area under the glucose concentration-time profile; 8121.4 and 8293.3 mg/dL*min for method 1-5 and method 6, respectively). The results call for a careful selection of the method for the assessment of insulin PK and PD parameters depending upon the objectives of study as the relationship between PK and PD could vary among methods and may lead to misinterpretation of data. Disclosure S.M. Chung: None. J.A. Penzenstadler: None. M. Khurana: None. C. Sahajwalla: None.