Seminar Details

Seminar Details


Friday

Feb 12

3:30 pm

The Bayesian Approach To Population Pharmacokinetic Modeling

Jon Wakefield (Joint with Biostatistics)

Seminar

Imperial College, School of Medicine - Department of Epidemiology and Public Health

Population pharmacokinetic/pharmacodynamic (PK/PD) analysis is an important part of drug development since it aids in the understanding of the complex relationship between dose, drug concentration and therapeutic response. Pharmacokinetics considers the absorption, distribution and elimination of a drug and its metabolites, and pharmacodynamics the action of a drug on the body. The aim of population PK/PD modeling is to identify sources of variability in concentrations and responses, and to quantify the remaining variability. In this talk I will illustrate how the form of the data, the specific aims, and the statistical analyses change during the various phases of drug development. Population PK/PD data are very naturally modeled within a hierarchical framework: within-individual variability is captured at the first stage and between-individual variability at the second stage. The first-stage individual-specific models are typically nonlinear and the dimensionality of the parameter space is large, and so inference is not straightforward. I outline a Bayesian approach to PK/PD modeling and illustrate the methodology in the context of three examples. First, I highlight parameterization issues and discuss population meta-analysis using concentration data obtained from healthy volunteers following administration of the anti-asthmatic drug fluticasone propionate. I address covariate modeling using concentration data obtained from new-born babies administered with the antibiotic vancomycin. The aim in this study was dosage recommendation. Finally, I demonstrate a specific application of errors-in-variables modeling in the context of population PK/PD analysis for the anti-coagulant recombinant hirudin. In this case the data consist of drug concentrations and clotting times from 301 patients who had undergone total hip replacement.