Development and application of a pharmacokinetic model for the glycation of albumin

Glycated haemoglobin, HbA1c, is used commonly as a marker for glycaemic control. In patients with chronic kidney disease (CKD) red blood cells (RBCs) are removed faster from the circulation, giving less time for glycation of haemoglobin to occur [1]. Thus, HbA1c concentrations are falsely lowered in these patients. Glycated albumin (GA) has been suggested as an alternative marker of glycaemic control in patients with CKD [2,3] since it is independent of the RBC lifespan and unaffected by CKD. Furthermore, GA has a half-life of only about 20 days and reflects changes in blood glucose concentrations faster than HbA1c.

The aim of the project was to develop a model that describes the time course of GA in order to assess its potential clinical benefits.

A model for GA was developed based on data extracted from the literature and modelled using NONMEM®. Simulations were carried out in MATLAB®. Predictions of the GA model were compared to predictions of clinical benefits from a model for the time course of HbA1c from [4]. GA and HbA1c model predictions were compared to investigate the difference in response to a change in 24h-mean blood glucose concentration when diabetes treatment commences.

The GA model described the literature data well. Between subject variability in GA was larger than for HbA1c. Simulation of a decrease in mean plasma glucose concentrations resulted in a faster change in GA compared to HbA1c.

GA could potentially be used as an alternative marker to assess glucose control in diabetic patients with CKD. GA could be used instead of HbA1c as biomarker in phase II clinical trials for novel antidiabetic drugs.

[1] Viljoen M et al. (1991). Nephron. 52(2):271-278.
[2] Inaba M et al. (2007). J. Am. Soc. Nephrol. 18(3):896-903.
[3] Peacock T et al. (2008). Kidney Int. 73(9):1062-1068.
[4] Kalicki R et al. (2009). PAGE 18 Abstr 1677 [], St. Petersburg, Russia.