N.J Rahman,J.C Wakefield,C.Falcoz,
Imperial College, Glaxo Wellcome
A retrospective pharmacokinetic (PK) analysis of data from a number of Phase I studies has been performed implementing the population approach with Bayesian methodology. Also, we evaluate what could have been a better design of the studies.
We first analysed data in the same individual obtained from a pivotal study which contained three doses. Each dose was analysed separately and then all three doses were analysed simultaneously after dose linearity had been established. The pivotal study was designed to establish the PK model with accuarate estimates of the PK parameters.
The data were analysed using Bayesian and Classical methods so that a comparison could be made. The software used were POPKAN and NONMEM.
Also available are earlier studies which consisted of lower doses. These studies were analysed to see if the model, established with the pivotal study, representing the concentration of the drug in the blood could be identified early enough to optimize the design of future studies. How the studies can be combined is also examined.
Reference: PAGE 5 (1996) Abstr 567 [www.page-meeting.org/?abstract=567]
Poster: poster