Subcutaneous insulin substitution in insulin-dependent diabetes mellitus. Pharmacokinetic and pharmacodynamic studies.

Determination of free and total insulin with radioimmunoassay, after precipitation of endogenous insulin antibodies with polyethylene glycol, was evaluated. Insulin substitution in insulin-dependent diabetic patients was investigated, embracing the 24 h free insulin and glucose profiles with different regimens, the miscibility of insulin preparations, the overnight metabolic control, and bolus doses of different size with infusion pumps. In the free and total insulin assay precipitation of immunoglobulins with polyethylene glycol was almost complete and the recovery was high. Compared to immediately precipitated and assayed plasma samples at 37 degrees C, free insulin slightly decreased in immediately processed serum (20 degrees C), and also in plasma after 3 h at 20 degrees C. In stored (-20 degrees C) unprecipitated plasma samples free insulin increased after 4 weeks and also in serum samples after 26 weeks, whereas stored PEG-supernates were stable. In healthy controls a low basal insulin was found, increasing about tenfold postprandially. No morning rise in free insulin or glucose was found. The 24 h free insulin profile was strikingly unphysiological with 1 or 2 dose regimens; there was preprandial and nocturnal hyperinsulinaemia but absence of meal-related free insulin peaks. A considerable glucose rise was found after breakfast. Intensive regimens with conventional injections or infusion pumps, gave 24 h free insulin profiles that were similar to the physiological. However, the prandial peaks were retarded; and hyperinsulinaemia was shown with infusion pumps during daytime. An immediate loss of regular insulin was demonstrated after mixture with semisynthetic human lente insulin in vitro and in vivo, but not after mixture with biosynthetic human NPH insulin. The morning glucose control was similar with a bedtime injection of intermediate-acting insulin or continuous subcutaneous insulin infusion, but less hyperinsulinaemia overnight was found with the infusion pump at a constant rate. No early morning glucose rise was demonstrated. Dose-related free insulin profiles were shown after bolus doses with an infusion pump, although they were retarded compared to the physiological postprandial response. The postprandial hyperinsulinaemia was aggravated by continuous subcutaneous insulin infusion. Glucose consumption during euglycaemic clamp corresponded to the free insulin profiles, indicating that free insulin represents the biologically active hormone.