Role of the Liver in Glucose Homeostasis Using Metabolic Imaging

The liver has a central role in maintaining glucose homeostasis. During periods following food intake the liver stores glucose whilst during fasting periods it produces and releases glucose into the circulation. These key regulatory features prevent hyperglycaemia after meals via increase in hepatic glucose uptake and prevent hypoglycaemia during food deprivation via hepatic glucose output. Although the exact numbers are unknown, it is suggested that approximately 25%-30% of an oral glucose load are taken up by the liver. Since hepatic glucose uptake is closely linked with hepatic glycogen synthesis, the fraction of an oral glucose load that is converted to glycogen is similar or somewhat less. Other pathways downstream of hepatic glucose uptake are the conversion to lactate, oxidation to carbon dioxide (CO2) or synthesis of fatty acids. Glycogenolysis and gluconeogenesis contribute to hepatic glucose output, in yet unknown proportions. Key regulators of hepatic glucose metabolism act through diverse mechanisms. Hepatic glucose uptake is mainly regulated by the level of insulin, the rate of glucose appearance in the portal vein, the portal-peripheral glucose and insulin gradient and neuronal signalling1. Hepatic glucose production is regulated by the provision of substrates such as lactate and glycerol, allosteric control by metabolites such as glucose, and balance of hormones such as insulin, glucagon and catecholamines. An imbalance between hepatic glucose uptake and hepatic glucose output results in dysglycaemia which can be both hyper- or hypoglycaemia.

Hepatic glucose metabolism is dysregulated in a broad spectrum of diseases. Prime examples are type 1 and type 2 diabetes in which altered hepatic glucose handling contributes to hyperglycaemia, although via distinct mechanisms. Whereas in type 2 diabetes, insulin resistance and hence impaired suppression of hepatic glucose output is the key pathophysiological feature, lack of the portal-peripheral insulin gradient (insulin levels normally threefold higher in portal vein than in arterial blood due to drainage of secreted endogenous insulin into the portal vein) seems to be more relevant in type 1 diabetes. In the latter case absolute insulin deficiency and hence coverage of total insulin requirements by the exogenous subcutaneous route generates a very different vascular insulin profile compared with endogenously secreted insulin. Experiments in conscious dogs showed that glucose uptake is equally divided between the liver and muscle when insulin is infused via the portal vein, but when insulin is delivered peripherally the percentage of glucose taken up by the liver is less than half of normal. These findings suggest that peripherally delivered insulin cannot replicate the physiologic regulation of postprandial hepatic glucose uptake, but direct evidence in humans is currently lacking.

Another condition that is characterised by an altered portal milieu are patients having undergone bariatric surgery. The re-arrangement of the gastrointestinal tract substantially alters the portal milieu by accelerated glucose fluxes and higher and earlier gut peptide hormone patterns. The two most commonly performed bariatric surgery procedures, namely Roux-en-Y gastric bypass, which re-routes the small intestine to a small stomach pouch, and sleeve gastrectomy, which reduces the stomach to about 15% of its original size, significantly accelerate glucose absorption. It was recently demonstrated that this effects is more pronounced after Roux-en-Y-gastric bypass than sleeve gastrectomy. Accelerated glucose absorption leads to higher glucose concentrations in the portal vein. Of note, animal experiments using portal vein catheterization showed that under elevated glucose levels in the portal vein promote hepatic glucose uptake, however direct evidence in post-bariatric surgery patients is lacking.

Organ-specific substrate exchanges (uptake and output) can be best studied by measuring arterio-venous substrate concentration difference and organ blood supply. The additional use of isotopically labelled substrates further allows calculating intra-organ turnover rate. Although invasive, this method can be applied for most organs or tissue, such as the kidney, heart, brain or whole limbs. The liver's anatomical location and connection to the portal circulation makes the the calculation of arterio-venous-substrate gradient in humans particularly challenging, however. Surgical catheterization of the portal vein in humans is not possible for practical and ethical reasons. As a consequence, current non-invasive approaches in humans rely on the use of stable isotopes and can only provide an estimate of splanchnic glucose uptake (sum of liver and intestinal glucose utilisation) but do not allow for the quantification of hepatic glucose uptake.

Since it is generally assumed that the liver is the sole source of glucose production (an assumption essentially verified in normal condition, since the kidney appears to contribute less than 10% total glucose output), a simplified tracer approach with analysis of the systemic dilution of infused labelled glucose can reliably estimate hepatic (endogenous) glucose output. However, such isotope dilution cannot estimate hepatic glucose uptake, which has essentially been indirectly assessed in multiple (oral+iv) glucose tracers experiment and calculation of the systemic appearance of ingested labelled glucose. These measurements are however tightly dependent on the mathematical model used and hence remain semiquantitative. Furthermore, they do not allow to differentiate gut and hepatic glucose uptake.

Thus, the only way to directly assess hepatic glucose uptake is through highly invasive portal vein catheterization which requires animal models. Such models can simulate postprandial hepatic glucose handling but applicability to humans are limited. Current concepts of hepatic glucose uptake under different conditions mainly stem from animal experiments in which overnight fasted conscious dogs underwent portal vein catheterization.

From the above mentioned dilemma it follows that obtaining quantitative data on hepatic glucose uptake in humans requires a non-invasive approach such as imaging. Positron emission tomography (PET) scanning with the tracer fluorine-18 (F-18) fluorodeoxyglucose (FDG), called FDG-PET enables direct observation of tissue glucose uptake by quantifying radioactivity over time in vivo. Some researchers have thus suggested to use FDG-PET to study human glucose metabolism. However, FDG-PET confers the major downside of exposing individuals to remarkable amounts of radiation, a risk that is not considered justified for research purposes only. In addition FDG-PET does not inform on metabolism downstream of glucose uptake and the intravenous administration route of the radioactive glucose is not reflective of normal physiology.

Clearly, there is a demand for a non-invasive, non-radioactive and easily applicable approach to investigate human hepatic glucose metabolism including the quantification of hepatic glucose uptake. Deuterium metabolic imaging is a novel, non-invasive imaging approach that combines deuterium magnetic resonance spectroscopic imaging with oral intake or intravenous infusion of nonradioactive 2H-labeled substrates to generate three-dimensional metabolic maps. Deuterium metabolic imaging can reveal glucose metabolism beyond mere uptake and can be used with other Deuterium (2H)-labeled substrates as well. It has recently been demonstrated by De Feyter et al. that deuterium metabolic imaging allows mapping of glucose metabolism in the brain and liver of animal models and human subjects using 6,6-2H2-glucose. Deuterium metabolic imaging is a promising, non-invasive and easy-to-implement imaging technique that opens new avenues to address important knowledge gaps such as the extent and dynamics of postprandial hepatic glucose uptake and utilisation.

Inclusion criteria

• Age≥18 years
• Capacity to give informed consent
• Willingness to adhere to the study protocol

In group II (type 1 diabetes), the following criteria must be met in addition:

• T1D for ≥2 years or evidence of undetectable C-peptide (<100pmol/l with concomitant plasma glucose≥4.0mmol/l)
• HbA1c≤8.0mmol/l (64mmol/mol)

In group III (bariatric surgery), the following criteria must be met in addition:

• Female
• Bariatric surgery (Roux-en-Y gastric bypass or sleeve gastrectomy) ≥1 year ago
• Lack of a history of diabetes or pre-diabetes (HbA1c≤5.6% in the absence of anaemia)

Exclusion criteria

• Pregnancy, planned pregnancy or breastfeeding
• Medication that interfere with glucose metabolism (participation requires discontinuation of these agents at least one week before the study visit) except for individuals with type 1 diabetes
• History of gastrointestinal surgery (other than bariatric surgery for group III)
• Known kidney, liver or heart disease
• Claustrophobia
• Contraindications to magnetic resonance imaging according to designated questionnaire
• Substance abuse
• Physical or psychological condition likely to interfere with the normal conduct of the study and interpretation of the study results as judged by the investigator