A model of hepatic glucose disposal using data from a labeled oral glucose tolerance test combined with Deuterium Metabolic Imaging

Mathematical modeling of glucose metabolism is a key topic in bioengineering, allowing to gain insights about the physiological mechanisms of glucose regulation and to improve clinical practice in the field of diabetes. So far, several models of whole-body glucose kinetics have been developed by exploiting plasma glucose and insulin measurements. However, despite the use of glucose tracers allowed the development of more sophisticated mathematical models, only a few models of glucose metabolism at organ-level have been developed due to experimental limitations. Deuterium Metabolic Imaging (DMI) is an emerging spectroscopic technique that has been recently used to measure glucose concentration in the liver, providing metabolic information at organ-level. The aim of this work is to develop a model of hepatic glucose disposal using DMI and peripheral glucose and hormones data after a labeled oral glucose tolerance test. Ten RYGB (2M, age=39.2±3.1 y, BMI=28±1 kg/m2 - mean±SE) subjects and 10 healthy controls (5M, age=36.4±3.2 y, BMI=26±1 kg/m2) ingested 60g of glucose (G) labeled with [6,6-2H2]-glucose. Plasma samples were frequently drawn to measure G, insulin (I), and tracer concentrations in the following 180 min, along with DMI measurements. A batch of models of hepatic glucose disposal were developed and identified at the individual level in both groups. The selected model was able to describe both plasma and hepatic glucose concentration data with the minimum number of physiologically plausible and precisely estimated parameters in both groups. In addition, model-derived estimated of rates of glucose appearance in the portal vein (u_HPV) and glucose disposal (R_d) were assessed and compared between the two groups. In particular, the model highlighted a significantly higher area under the curve (AUC) of u_HPV and R_d in RYGB vs. control, both after 60 min from glucose intake and at the end of the experiment, while no differences were observed in disposal glucose effectiveness (GE^D) and disposal sensitivity to insulin (S_I^D), showing that the known faster glucose kinetics characterizing the post-operative group is mainly due to the gastrointestinal alterations after surgery. In addition, the model was also validated against a state-of-art model of whole-body glucose kinetics showing a statistically significant correlation for GE^D, S_I^D, and AUCs of systemic rate of appearance and disposal fluxes at 60 and 180 min. Future work will aim to expand the model structure to describe endogenous glucose kinetics, possibly exploiting nonlinear models of transport and disposal fluxes and accounting for the direct effect of insulin on hepatic disposal. Lastly, a model of glycogen metabolism will be considered to complement the present model with information regarding the specific pathways of hepatic glucose disposal.