Main glucose hepatic fluxes in healthy subjects predicted from a phenomenological-based model

• Published in: Computers in biology and medicine Vol. 142; p. 105232
• Main Authors: Builes-Montaño, Carlos E., Lema-Perez, Laura, Garcia-Tirado, Jose, Alvarez, Hernan
• Format: Journal Article
• Language: English
• Published: United States Elsevier Ltd 01.03.2022; Elsevier Limited
• Subjects: Blood glucose; Blood Glucose - metabolism; Carbohydrates; Computer simulation; Diabetes; Diabetes mellitus; Diabetes Mellitus, Type 2; Enzymes; Fasting; Fatty acids; Fluxes; Glucagon; Glucagon - metabolism; Gluconeogenesis; Glucose; Glucose - metabolism; Glucose tolerance; Goodness of fit; Healthy Volunteers; Hepatic vein; Homeostasis; Humans; Insulin; Insulin - metabolism; Internal Medicine; Liver; Liver - metabolism; Mathematical model; Mathematical models; Metabolism; Other; Pancreas; Parameters; Regulation; Veins & arteries; Glucose metabolism; Mathematical model; Diabetes mellitus; Liver
• ISSN: 0010-4825; 1879-0534; 1879-0534
• DOI: 10.1016/j.compbiomed.2022.105232

The liver has a unique role in blood glucose regulation in postprandial, postabsorptive, and fasting states. In the context of diabetes technology, current maximal models of glucose homeostasis lack a proper dynamical description of main glucose-related fluxes acting over and from the liver, providing a rather simplistic estimation of key quantities as endogenous glucose production and insulin and glucagon clearance. Using a three-phase well-established phenomenological-based semi-physical modeling (PBSM) methodology, we built a detailed physiological model of hepatic glucose metabolism, including glucose utilization, endogenous glucose production through gluconeogenesis and glycogenolysis, and insulin and glucagon clearance. Mean absolute errors (MAE) were used to assess the goodness of fit of the proposed model against the data from three different in-vivo experiments -two oral glucose tolerance tests (OGTT) and a mixed meal challenge following overnight fasting-in healthy subjects. Needing little parameter calibration, the proposed model predicts experimental systemic glucose mean ± std 5.4 ± 5.2, 7.5 ± 6.8, and 7.5 ± 7.5 mg/dL, in all three experiments. Low MAEs were also obtained for insulin and glucagon at the hepatic vein. The quantitative concordance of our model to the experimental data exhibits a potential for its use in the physiological study of glucose liver metabolism. The model structure and parameter interpretability allow the union with other semi-physical models for a better understanding of whole-body glucose homeostasis and its use in developing diabetes technology tools. •A novel multiparametric physiological model of the liver's role in glucose metabolism.•Minimal parameter calibration to replicate results of diverse in-vivo experiments.•This model could be used in digital platforms to aid precision medicine in diabetes.•Quantitation of glucose, insulin, and glucagon concentrations at the hepatic vein.•High potential for in-silico experimentation to recreate glucose metabolism in humans.