Moustafa M.A. Ibrahim (1), Anna Largajolli (1), Mats O. Karlsson (1), Maria C. Kjellsson (1)
(1) Department of Pharmaceutical Bioscience, Uppsala University
Background and Objectives: An Integrated minimal model (IMM) has been proposed by Largajolli et al. to characterize simultaneously glucose-insulin regulation system following intravenous glucose tolerance test (IVGTT) [1]. This model was developed using the population nonlinear mixed effect modelling approach. It provides less-invasive estimates of insulin sensitivity index (SI) and glucose effectiveness index (GE) with full simulation capabilities. Since this model was developed with unlabelled IVGTT data, it cannot separate endogenous glucose production from glucose disposal, leading to a biased parametric description of glucose disposal, particularly GE reflect both mass action and control mechanism, and it overestimates fractional glucose clearance [2]. Here, we investigated the presence of a biased parametric description of glucose disposal and its impact on the key parameters for clinical diagnosis, glucose effectiveness and insulin sensitivity, by simultaneous fitting of cold and hot glucose from labelled IVGTT. We investigated as well the use of prior functionality to analyze unlabeled IVGTT data to obtain refined IMM parameter estimates.
Methods: Data was obtained from two previously published studies of stable isotopically labelled intravenous glucose tolerance test (IVGTT) performed in healthy subjects with one of the studies being insulin-modified [3, 4]. Frequent plasma samples were analysed to determine the concentration of total glucose, insulin, and hot glucose. The structure of the hot glucose disposition model was assumed to be identical to that of unlabelled glucose. Informative priors were obtained from one of the studies and the IMM was fitted to the other study using prior functionality in NONMEM and only total glucose observations.
Results: For the glucose sub-model, IMM parameter estimates significantly differed from IMM parameter estimates, in particular for GE which was estimated to be twice as high in the absence of hot glucose data. VPCs from these two models were similar albeit the different parameter estimates. Parameters of insulin sub-model were less affected and insulin sensitivity could be drived from unlabelled data. The use of priors did not provide a solution to the biased estimates of glucose disposal.
Conclusion: Endogenous glucose production in IMM is poorly described, leading to the biased description of glucose disposal. Although incorporation of hot glucose will not change the fact that EGP is constant, but it will provide a more reliable description of glucose metabolism during an IVGTT and provide less biased estimate of glucose effectiveness.
Absence of Hot glucose data will mislead IMM use to explore diabetes and drug effects.
Our analysis suggests that priors do not offer a suitable alternative in absence of labelled glucose data. However the priors came from a study with a different design than the analyzed (insulin-modified) which may contribute to the poor fit with priors
References:
[1] Largajolli A., Bertoldo A.,Cobelli C., and Denti P., An integrated glucose-insulin minimal model for IVGTT, PAGE 22 (2013) Abstr 2762 [www.page-meeting.org/?abstract=2762]
[2] Avogaro, A., J. D. Bristow, D. M. Bier, C. Cobelli, and G.Toffolo. Stable-label intravenous glucose tolerance test minimal model. Diabetes 38: 10481055, 1989.
[3] Vicini P, Caumo A, Cobelli C. The hot IVGTT two-compartment minimal model: indexes of glucose effectiveness and insulin sensitivity.Am J Physiol. 1997; 273:E1024-E1032.
[4] Vicini P, Zachwieja JJ, Yarasheski KE, Bier DM, Caumo A, Cobelli C. Glucose production during an IVGTT by deconvolution: validation with the tracer-to-tracee clamp technique. Am JPhysiol. 1999; 276:E285-E294
Reference: PAGE 25 (2016) Abstr 6007 [www.page-meeting.org/?abstract=6007]
Poster: Drug/Disease modeling - Endocrine