Pharmacokinetic-pharmacodynamic model for the anticholinergic effect of glycopyrrolate

Objective: The purpose of this study was to develop and test a pharmacokine tic-pharmacodynamic (PK-PD) model for the anticholinergic effect of glycopy rrolate in eight healthy male volunteers. Methods: First, arterial drug concentration (C-p) data after a single intra venous (i.v.) bolus injection (5 mug/kg) were individually fitted to a thre e-compartment PK model. Second, the effect of a 2-h glycopyrrolate i.v. inf usion (5 mug/kg/h) on the mean R-R interval (RRI) and the Hayano index of t he high frequency variability of RRI (HF CCV) was modelled using an effect- compartment, inhibitory sigmoidal E-max model, with the individual PK param eters from the first part as constants. Third, the developed model was test ed using a computer-driven infusion which aimed at two ascending steady-sta te effect-site concentrations (C-p) at l-h intervals, corresponding to 20% and 80% of the maximal effect (E-max) observed in the second part. Results: Modelling of the HF CCV data yielded the following mean(+/- SD) es timates: concentration at 50% of E-max (EC50), 2.46 +/-0.58 ng/ml, equilibr ation half-time (t(1/2) (ke0)), 42.5 +/- 7.7 min, and sigmoidicity factor ( gamma), 7.26 +/- 2.82. The corresponding values for RRI data were 2.79 +/- 0.52 ng/ml, 58.3 +/- 17.2 min, and 4.75 +/-1.56. During the computer-contro lled two-step infusion (performed using HF CCV as the effect variable), the measured C-p approached the targeted C-p in most of the subjects, while th e observed effect appeared to surpass the targeted levels. Conclusion: Although we were able to develop individual PK-PD models for gl ycopyrrolate, maintaining a stable anticholinergic effect in the computer-d riven infusion appeared to be difficult. This is probably due to intraindiv idual variability in the PK-PD parameters and the extremely steep concentra tion-effect relationship of glycopyrrolate.