Comparison of intraperitoneal, intraportal and intravenous insulin infusion.

In order to avoid complications induced by long-term infusion of insulin into the portal vein, we examined the effect of intraperitoneal (ip) insulin infusion on arterial plasma insulin and glucose concentrations in 6 pigs, made diabetic by a constant intravenous (iv) infusion of glucose, epinephrine and propranolol. Insulin was infused by an electromechanical programmable mini-pump (Pharmaject Micro Infusion System, Pharmacia Electronics) as a booster injection of 46 mU highly purified porcine insulin Leo/kg body weight, followed by 3 infusion periods of 30 min each with stepwise decreasing rates of 1.6--0.8 and 0.2 mU/kg/min in a total volume of 192 microliters. Insulin was infused in a peripheral vein, a portal vein and into the peritoneal cavity. A steep rise of arterial plasma insulin was demonstrated followed by a slow and identical decline in the peripheral and portal experiments, whereas only a small increase of plasma insulin was seen in the ip experiment, indicating insufficient absorption of insulin from the peritoneal cavity. The decrease of plasma glucose was identical in the peripheral and portal vein experiments, indicating that insulin infused in the portal vein does not seem to have a higher hypoglycaemic effect, than insulin infused in a peripheral vein. Intraperitoneal insulin infusion seems not to be a practical substitute for iv insulin infusion.