Féry, F., L. Tappy, J. Devière, and E. O. Balasse. Comparison of intraduodenal and intravenous glucose metabolism under clamp conditions in humans. Am J Physiol Endocrinol Metab 286: To determine whether the uptake and metabolic partition of glucose are influenced by its delivery route, 12 normal volunteers underwent two 3-h euglycemic (ϳ93 mg/dl) hyperinsulinemic (ϳ43 mU/l) clamps at a 3-to 5-wk interval, one with intravenous (IV) and the other with intraduodenal (ID) glucose labeled with [3-3

more »

... H]-and [U-14 C]glucose. Systemic glucose was traced with [6,6-2 H2]glucose in eight subjects. During the last hour of the clamps, the average glucose infusion rate (5.85 Ϯ 0.37 vs. 5.43 Ϯ 0.43 mg⅐kg Ϫ1 ⅐min Ϫ1 ; P ϭ 0.02) and exogenous glucose uptake (5.66 Ϯ 0.37 vs. 5.26 Ϯ 0.41 mg⅐kg Ϫ1 ⅐min Ϫ1 ; P ϭ 0.04) were borderline higher in the ID than in the IV studies. The increased uptake was entirely accounted for by increased glycolysis ( 3 H2O production), which was attributed to the stimulation of gut metabolism by the absorptive process. No difference was observed in glucose storage whether it was calculated as glucose uptake minus glycolysis (ID vs. IV: 2.44 Ϯ 0.28 vs. 2.40 Ϯ 0.31 mg⅐kg Ϫ1 ⅐min Ϫ1 ) or as glucose uptake minus net glucose oxidation (2.86 Ϯ 0.33 vs. 2.81 Ϯ 0.35 mg⅐kg Ϫ1 ⅐min Ϫ1 ). Because peripheral tissues were exposed to identical glucose, insulin, and free fatty acid levels under the two experimental conditions, we assumed that their glucose uptake and storage were similar during the two tests. We therefore suggest that hepatic glycogen storage (estimated as whole body minus peripheral storage) was also unaffected by the route of glucose delivery. On the other hand, in the ID tests, the glucose splanchnic extraction ratio calculated by the dual-isotope technique averaged 4.9 Ϯ 2.3%, which is close to the figures published for IV glucose. Despite the limitations related to whole body measurements, these two sets of data do not support the idea that enteral glucose stimulates hepatic uptake more efficiently than IV glucose. hepatic glycogen; indirect colorimetry; glycosis; portal signal THE OLD AND ATTRACTIVE CONCEPT that the liver, owing to its unique position between the portal vein and systemic circulation, functions as a "filter" that retains significant amounts of glucose absorbed by the gut has been challenged for many years. A simple and remarkable experiment conducted more than 30 years ago has greatly contributed to minimize the importance of this "anatomical privilege" of the liver in relation to postprandial glucose metabolism. McIntyre et al. (27) indeed showed in the dog that the infusion of equal amounts of glucose into the portal vein and a systemic vein produced identical peripheral levels of both glucose and insulin. This result does not in itself exclude the possibility that hepatic uptake and/or storage of glucose might have been stimulated to a greater extent during the infusion via the portal vein. This possibility is supported by the results of several subsequent animal studies (7, 18, 21, 37), but not all (2, 6, 30), in which hepatic (splanchnic) glucose metabolism was compared after oral (or intraportal) and intravenous (IV) administration. To investigate the possible mechanisms of such a stimulation in the dog, Cherrington and coworkers (see references in Ref. 8) carefully examined the interplay between the various factors controlling hepatic glucose uptake and storage. They observed that, in addition to stimulation by insulin and by the glucose load reaching the liver, hepatic glucose uptake and glycogen synthesis are also stimulated by a "portal signal" activated by a positive portal-arterial glucose concentration gradient. Although the exact nature and location of this putative sensor are not yet known, it seems to act through local neural mechanisms (1). It has been proposed that this signal allows the organism to distinguish between meal-derived and endogenously produced (or peripherally administered) glucose so as to direct the former preferentially to the liver. In one animal study (20), it was suggested that splanchnic glucose delivery results in the stimulation of net splanchnic glucose uptake whether the glucose entered the liver via the hepatic artery or via the portal vein. In humans, relatively few authors have compared the splanchnic metabolism of oral vs. intravenous glucose. Using the hepatic vein catheterization technique, DeFronzo et al. (11) showed that, under hyperglycemic clamp conditions, glucose ingestion stimulates net splanchnic glucose uptake by a factor of six. The higher insulin levels that prevailed after glucose ingestion and the possibility that the oral load might have been incompletely absorbed within the 3 h of the study might have led to an overestimation of these results. However, on the whole, these data are clearly in favor of the operation of a portal signal in humans. On the other hand, using a dualisotope technique, Radziuk (31, 32) found no difference between the intravenous and intraduodenal routes regarding hepatic glycogen storage. In this study, performed under nonsteady-state conditions, the intravenous glucose was infused at a rate that matched the rate of absorption of an oral glucose load, and no attempt was made to control glycemic or insulinemic excursions. To further explore the impact of the route of administration on glucose metabolism in humans, we recently compared the Address for reprint requests and other correspondence: