between-patient variability in PK and PD

From: "R.A.A. Mathot" r.mathot@erasmusmc.nl Subject: [NMusers] between-patient variability in PK and PD Date: Wed, May 4, 2005 11:08 am Dear all, Currently, I am working on a PK-PD dataset of 250 patients. In the study each patient received 4 different drug doses and effects were recorded. PK were linear; between- and within-patient variability in clearance were 35% and 20%, respectively. The relationship between drug effect and AUC was described according to a sigmoidal Emax model. Between-patient variability in AUC50 was large: ca 150%. Since between-patient variability in PD is much larger than in PK, I am wondering what the assessment of PK contributes to the evaluation of the dose-respons relationships of this drug in this study. In fact more or less comparable sigmoidal relationships were obtained when drug effect was related to the administered dose: between-patient variability in ED50 was 130%; the objective function was 50 points higher than for the AUC-effect relationship indicating a worse fit; residual error was similar. I am wondering whether there are some objective criteria which may be used to decide whether it is useful to include PK in future studies of the dose-respons relationship of this drug. I am looking forward to your comments. ======================================================R.A.A. Matht Apotheek - Laboratorium Erasmus MC