Khan, A.Z.1, Pitsiu, M.1, Aarons, L.A2, Parivar, K.3, Widman, M3
1 Medeval Ltd, University of Manchester, Manchester Science Park, Manchester, England. 2 Dept. of Pharmacy, University of Manchester, Manchester, England, 3 Pharmacia Upjohn, Stockholm, Sweden.
Eltanolone, 3a-hydroxy-5b-pregnan-20-one, is a naturally occurring metabolite of progesterone, which has been under development for use in humans for the induction of anaesthesia.
The objectives of this analysis were to define a population pharmacokinetic model for eltanolone, having designed a prospective sparse data study in patients.
Initial analysis was carried out on data available from phase I volunteer studies in order to confirm the structural model (three compartment, iv administration). The outcome of this analysis was used to identify optimal sampling windows (D-optimisation) for the prospective population study.
The final analysis was performed on a combined data set, consisting of sparse data (230 patients) and rich data (10 patients). The inclusion of the rich data proved to be necessary for the stabilisation / convergence of the algorithm. Covariate relationships (gender, age, disease state, therapy, weight, height) were investigated using multiple linear regression (SAS) and subsequently included in the NONMEM model.
In addition to successfully identifying the structural and variance model parameters, weight was the single best estimator of eltanolone clearance.
The model was applied to a paediatric population and predictions were compared with the available observations.
Reference: PAGE 6 (1997) Abstr 659 [www.page-meeting.org/?abstract=659]
Poster: poster