Pharmacokinetic-pharmacodynamic modelling of human insulin: validity of pharmacological availability as a substitute for extent of bioavailability.

A method for assessing the extent of bioavailability (EBA) of human insulin from pharmacological data was determined. The time course governing increases in the plasma concentration of immuno-reactive insulin (IRI), as well as its pharmacological effects (glucodynamics), was determined after the intravascular administration of varying doses of human insulin. Pharmacokinetic (PK), pharmacodynamic (PD), and link models were constructed to elucidate the quantitative relationship between plasma IRI levels and pharmacological effects. After extravascular administration of the test formulation, only the time course governing the observed pharmacological effects was determined. The pharmacological data was translated into theoretical plasma concentration data, using the PK-PD model. Following this, the area under the theoretical plasma concentration-time curve (AUC) of the test formulation was calculated. The EBA was then estimated against a reference (intravascular) formulation, using a conventional equation. Since the pharmacological effects of insulin were observed to be highly dosing-rate-dependent, the PD model used in this study was modified to apply over a wide range of infusion rates. The results of the PK-PD analysis indicate that the doses administered can be accurately predicted from pharmacological data. To validate this method, the EBAs of controlled release formulations (the Osmotic Mini Pumps) of insulin, subcutaneously administered to the rat, were estimated. The EBA values obtained (92-96%) fell within a reasonable range. The area under the effect-time curves (AUE) obtained following subcutaneous applications of the Osmotic Mini Pump were calculated in a model-independent manner, in addition to pharmacological availabilities (PA), which were estimated against the reference (intravascular) formulations. The estimated PA values varied from 312% to 78%, in accordance with the intravascular input rates of the reference formulations. This indicates that PA should not be used as a substitute for EBA, unless data involving similar intravascular dosing rates to that of the reference formulations is available.