Effect of Intraduodenal Glucose Administration

Received for publication 24 August 1972 and in revised form 2 April 1973. INTRODUCTION The anatomic location of the liver, receiving the entire pancreatic output of insulin prior to its entry into the systemic circulation, and its ability to degrade insulin permits it to play an important role in regulating peripheral insulin concentrations (1). Although increases in peripheral insulin levels have been assumed to reflect increased pancreatic secretion of insulin, Stern, Farquhar, Silvers, and Reaven (2) emphasized the paucity of experimental evidence to support this assumption. They pointed out that their measurements of the rate of delivery of insulin into the general circulation in humans was not necessarily equivalent to the total insulin secreted by the pancreas. They also criticized the concept that the metabolic clearance rate of insulin measured insulin secretion because it does not include that fraction of the insulin secreted by the pancreas which is removed by the liver prior to its entry into the general circulation. Wright (3) also indicated that hepatic extraction of insulin could influence significantly peripheral insulin concentrations. Hepatic extraction of insulin has been measured by several different techniques. Mortimore, Tietze, and Stetten reported that the isolated perfused rat liver removed 40% of ["I] insulin (25 mU/ml) in a single passage (4). Insulin uptake was significantly less during perfusion of the hind limb. Madison and Kaplan estimated that approximately 54% of the 0.65 U of insulin injected into the portal vein of anesthetized patients was extracted by the liver during a single transhepatic passage (5, 6). Insulin extraction in these studies was evaluated using trichloroacetic acid precipitable radioactivity. However, Marshall, Gingerich, and Wright have emphasized the hazards of using [PI] insulin and measurement of trichloroacetic acid precipitable radioactivity as an indication of insulin metabolism (7). Samols and Ryder measured insulin concentrations by immunoassay in the hepatic vein and a systemic artery in patients with portal-caval The Journal of Clinical Investigation Volume 52 August 1973 -2016-2028 2016 shunts or thrombosis of the portal vein (8). Estimation of hepatic blood flow permitted calculations of insulin flux across the liver but the results may have been influenced by the liver disease. In four such patients, hepatic extraction averaged 41% of the 0.2 mU/kg per min endogenous insulin. Although these results indicated basal hepatic extraction of approximately 50% of the insulin presented, conflicting data exists concerning hepatic extraction of insulin after administration of glucose. Kaplan and Madison reported that glucose decreased the fraction of insulin presented to the liver which was removed by that organ (6). Infusion of 9 g of glucose over 17 min decreased hepatic extraction of insulin from an average of 54 to 38%. When 52 g of glucose was infused over 165 min, only 7.8% of the administered [1I]insulin was removed by the liver during the initial passage. However, the studies of Waddell and Sussman suggested contrary findings (9). In unanesthetized dogs, glucose administration increased the discrepancy between portal and peripheral vein insulin concentrations. When the portal vein insulin concentrations were above 500 AU/ ml, almost all of the insulin was removed by the liver in contrast to when the portal vein insulin concentration was below 50 uU/ml when almost none was extracted. However, since appropriate blood flows were not measured, hepatic extraction of insulin could not be quantitated. The present studies were undertaken to examine hepatic extraction of insulin in a more quantitative fashion and to determine whether this process is modified by glucose administration into the duodenum with its resultant increase in pancreatic insulin secretion. The results confirm that about 50% of the insulin presented to the liver in the basal state is removed during a single passage; following glucose administration there is no decrease in the percent of insulin removed by that organ, and some evidence to suggest a transient increase.