Insulin and Glucagon Breakthrough of Somatostatin Suppression: Importance of Portal Vein Hormone Measurements

Abstract

The infusion of somatostatin (800 ng/kg/min) into anesthetized dogs leads to suppression of portal insulin concentrations from 58 ± 5 μU/ml to 18 ± 5 μU/ml and peripheral arterial concentrations from 18 ± 2 μU/ml to 8 ± 2 μU/ml, these levels remaining suppressed for the duration of the infusion (210 min). Additional experiments were performed during which somatostatin (800 ng/kg/min) was infused alone for 30 min and then together with glucose to reach steady-state glucose concentrations of 422 ± 8 mg/dl. This hyperglycemic stimulus resulted in progressive increments in portal insulin of 15 ± 3, 22 ± 8, 41 ± 11, 62 ± 14, 77 ± 17, and 88 ± 6 μU/ml at successive 10-min intervals. The corresponding increments in arterial insulin were much less, being 5 ± 2, 4 ± 1, 5 ± 2, 8 ± 2,10 ± 4, and 19 ± 6 μU/ml. In separate experiments, the rate of glucose infusion was varied so that steady-state glucose concentrations of 173 ± 3, 255 ± 8, and 359 ± 5 mg/dl were achieved in successive 60-min periods. Portal vein insulin values in each period were 100 ± 34, 262 ±117, and 596 ± 162 μU/ml, with corresponding peripheral levels of 12 ± 6, 26 ± 14, and 69 ± 45 μU/ml, respectively. Similar experiments in which a lower dose of somatostatin was used (200 ng/kg/min), in the face of similar steady-state glucose increments, resulted in portal insulin levels of 182 ± 62, 435 ± 138, and 512 ± 102 μU/ml, and corresponding peripheral levels of 34 ± 15,141 ± 55, and 188 ± 20 μU/ml, respectively. Somatostatin infusion (800 ng/kg/min) suppressed portal vein glucagon levels to 101 ± 16 and peripheral levels to 39 ± 3 pg/ml. In these animals, hypoglycemia was then induced by infusion of insulin, and portal and peripheral glucagon levels rose to 341 ± 11 and 146 ± 27 pg/ml, respectively. Thus, the inhibition of insuiin and glucagon secretion by pharmacologic doses of somatostatin may be overcome by an appropriate stimulus. Because of the large capacity of the liver for insulin extraction, increases in portal insulin levels may not be reflected in the periphery. However, increases in portal vein glucagon, although underestimated from peripheral levels, were always associated with changes in peripheral levels, because the liver extracts less glucagon than insulin. These findings should be considered in the interpretation of experimental studies in which somatostatin has been used to suppress endogenous pancreatic hormone secretion and peripheral hormone levels alone are used to monitor the adequacy of suppression.